Circadian science Interviews Technology

Interview With Bharath A.

Thanks so much for joining us this morning! Could you tell us a little bit about yourself and your research interests?

I am currently a theoretical and computational chronobiologist. I work in the area of circadian clocks. My PhD was in the field of wireless communications with a focus on designing new technology to provide higher rates of data transfer between phones and data stations. In a sense, cellular technology and chronobiology are connected because your cellphone’s functions depend heavily on oscillators. There is a lot to do with oscillators, amplitude, and phases, which transitions into circadian clocks and circadian rhythms. After completing my PhD in electrical engineering in the US, I relocated to Germany and explored my interest in biology, ultimately combining my knowledge of biology and engineering at a theoretical biology institute. Now, I have my one small group here at Humboldt University in Berlin.

You’ve done a lot—from blood tests for circadian rhythms to pure modeling work. What is the direction you’re most excited about in your current research? 

I am most interested in circadian medicine, (or chronomedicine). We began looking at how rhythms were generated, starting with the negative feedback loop, which was awarded the Nobel prize in 2017. In the late ‘90s and early 2000s, all the clock genes were found, and then we moved into using kinds of massive sequencing technology. We started to look at initial clock outputs, which include all the transcriptomes, and peripheral clocks in all the different tissues.

Through proteomics and metabolomics, we now have a very good idea of how rhythms and clock outputs are generated. We have also started looking at how these outputs couple with different physiological processes and how they mediate the interaction between clocks, immunity, metabolism etc. It really does seem to affect almost everything. So, I think the next natural step is to look at how we can leverage this understanding to help people improve their health. I am taking the more computational and theoretical angle, including blood tests, towards improving people’s health through circadian medicine. 

I am also involved in clinical studies that explore what has already been done in mice, although human clocks, of course, are more heterogeneous. I tend to focus on the molecular outputs of the human clock. We are trying to map out how different clocks look in peripheral tissues of specific patient groups or subpopulations. We are also currently trying to associate these data with other clinical parameters. 

I started at Arcascope about a year ago, and that’s when I learned about circadian rhythms and how it affects you day-to-day. I’m still learning, but I find the field of chronomedicine very exciting. 

What we are talking about is cutting edge in terms of medication. Even simple things, such as timing—whether to consume medicine during the day or night—have a huge impact. For example, my colleagues discovered this in a patient’s case of rheumatoid arthritis. 

The symptoms of rheumatoid arthritis are strongest early in the morning, with certain markers ascending really early in the morning, around 4:00 or 5:00 AM. So they took the standard medication, formulated a delayed-release form, and asked patients to take it at night so that the medication was released when the disease symptoms were at their highest rate. It was a very, very simple chrono-solution. I think it was published in Lancet, around eight or ten years ago in the early days of chronomedicine. Some very simple things, such as timing medication, can improve the efficacy of medicines that already exist. 

How has your work on biological rhythms changed the way you live your life day to day?

I will divide this into two parts of my life: before kids and after kids. After having kids and experiencing the coronavirus pandemic, things have become more chaotic. I have always had a fairly regular circadian lifestyle. Most of my life, I could survive with maybe six or seven hours of regular-quality sleep. I would go to bed at 9:00 PM and wake up at around 5:30 AM. And only after University, I started staying up later, typically going to bed at 11:00 PM and waking up in the morning around 6:00 AM.

I never slept in much because I was always awake once the sun was out. And now with kids, I try to work when they are asleep. So I end up staying up a lot later than I should. I know that I should not eat at night, but I still do it even though I know it’s wrong, so I will try to change that.

In fact, the funny thing was that after we created this blood test, we did the blood test on ourselves. At the time when my son was an infant, we had really crazy sleep/wake schedules. Still, when I did the blood test to determine my chronotype, it was actually the most consistent measurement of all the people we had tested. We tested at three different days and my chronotype estimate was within ten minutes of each other. So, objectively my chronotype looks pretty stable in spite of changing seasons and having to endure a crazy sleep/wake schedule.

Olivia, Arcascope’s CEO, is actually working on a blog post about how sleep regularity could be more important than sleep duration. 

Yes, possibly. As a chronologist, we do not really know how long it takes before one starts feeling the negative effects of an irregular sleep schedule. At the moment, I’m not living a chronologically good lifestyle based on what we know. At a recent meeting, we had a discussion on the effects of living against your clock and it is still unclear. 

So we talk about correlations, which is all we have. We do not really know whether it’s a consequence, which way the causality goes or whether it is bidirectional. So, it is hard to say. Essentially if we change the balance, we change the likelihood that something can go wrong.

For example, Satchin Panda, the author of The Circadian Code, discusses fasting, and emphasizes that a 12-hour fasting period at night is very important. That is one very simple takeaway from his work. I tried fasting for the longest time but it is hard, especially if you stay up quite late after dinner, then it’s hard to fast. I realize my kids can fast because they end up sleeping for 12 hours. Although, I personally know scientists in his lab who have been regularly fasting for quite some time.

And that’s an interesting thought too, is if you do eat outside of your optimal meal window, what exactly does that do?

At this point, we do not know. I do know about [timing and] caloric restriction from Joe Takahashi. It is not just about when you eat, but also how much you eat. Also, if you have a long fasting window, you probably eat less throughout the day; on average, you eat less if you have a longer fasting period. This means that for the larger part of your wake time you are not eating.

As the former public outreach fellow for SRBR, what did you learn from the experience?

The biggest takeaway is that this field, initially narrow, is now really broad. We focus on how our rhythms are generated and the outputs of the clock. Now, since we are at the point where we are looking at interactions with other systems, you can pick any other aspect of physiology and apply a clock angle. There are people working on that interface, looking at immune clocks, microbiome clocks, cancer clocks, metabolism clocks and endocrinology clocks. 

From a personal perspective, I think that this broad field is challenging because I am not a biologist with the background information that one needs to understand a lot of this. I have to know about clocks, but I have to understand some background in these other fields as well, and being in this position forced me to keep abreast of everything in these interfaces.

Through networking, I met a lot of people. I was the fellow at the beginning of the coronavirus pandemic, so I was networking while everything was digital, including our SRBR conference, which was hosted online in 2020.

It was actually surprising, people didn’t really have any baseline ideas about how this was going to work. And it worked quite well. At that point, people were quite wired and they were kind of happy to put up with anything to still function in the sciences.

One of the rewarding things, which I found out after the fact, is that a lot of people found the content produced from SRBR, which helped people create or maintain some sort of connection with the outside world since they could not travel to meetings. Everybody was more active; there were a lot of PhD students and postdocs.

How do you think we can communicate our results better as a field in general?

This is something that is close to my heart. As an SRBR public outreach fellow, my true goal was to do public outreach, although we also did a lot of “in-reach”. We did a lot of outreach within the chrono-community, which is important to other scientists who are less aware of chronobiology, similar to when you entered the industry and you were less aware of how much clocks impact other fields. 

One example was the Daylight Savings Time (DST) laws during my time as the public outreach fellow. We had a town hall and still, in spite of so much science being out there, we did not get the result that we wanted. In a sense, we got the exact opposite result. Of course, that is not purely because of the science, but communicating science is also important from that perspective. As scientists, communication is also important, since the science must be seen in the context of lifestyle, society and general health. I think we have to become better at talking to all kinds of different people. 

Sometimes that comes with being less preachy about the sciences. In the DST example, we realize that people may have other reasons for wanting something different. Everything is not just about the things that we, as chronologists, care about.

And then there are methods of communicating science. Twitter is now becoming a standard tool for communicating science at the technical level, and we are quite active on Twitter. Nowadays, I think Twitter, at least in the chronobiology field, is here to stay. If we want to reach other audiences, we obviously have to go beyond just Twitter, such as videos and visuals on TikTok.

I struggle with using videos and visuals to tell a story about chronobiology. Our research is publicly funded: the taxpayer basically funds our research. This means that we have an obligation to report our findings back to the public and help them understand and lead better lives. Scientists, who are good at writing papers, have to communicate better with the public by creating content for social media, such as this very interview, but generating content takes a lot of time and effort. 

And of course, scientists are busy people who have lots of tasks on their plates. There is a group of people who are convinced that the pluses outweigh the minuses, and that they should generate content. But then there are people who would like to generate content, but they just do not have any bandwidth or they do not really know how to do it. That’s the bottleneck, really.

I think postdocs and PhD students are much more familiar with these tools. Some of them are creative and they already generate content. Actually, during SRBR, we did this trainee session on why public outreach is important on social media and how people can get involved. 

We asked others in the chrono-community, “How has being on social media helped your science?” This question presented an interesting opportunity to network among a community of friends who maintained connections and continued to ask questions during the pandemic.

So it sounds like there are a lot of pros to being online.

One must concede that while social media is great, there are some negatives and some people do not use social media. Just talking to an audience on social media platforms is not sufficient, and one has to look at other ways to reach other groups of people, such as speaking in different spaces. If you are senior, maybe you can get interviewed by a magazine or a newspaper, for example, the New York Times. 

So a recent example is John Hogenesch’s article, which was published in the New York Times about a month ago. It is a whole write-up on health and circadian medicine, arguing the case for why we should do it.

I wanted to take the last few minutes to ask if you had any work of yours that we haven’t touched on yet that you would like to highlight.

One recent controversial (not so in my opinion) practice is the use of Venn diagrams in circadian biology studies because it tends to overestimate the impact of all interventions on the circadian clock. This study was published last year which has had a very broad impact, not just in chronobiology but in all kinds of omics studies. One of my passions is teaching. I host trainee sessions where I teach people about the analysis of circadian rhythms, and have taught at two summer schools and three trainee days. I do this regularly to help people analyze circadian rhythms, and in the end, do the science correctly.

Circadian science Technology

Visualizing MESA: Part 2

We’ve already looked at the Multi-Ethnic Study of Atherosclerosis (MESA) dataset—an absolute treasure trove of sleep data, available from the NSRR at sleepdata.org—once, through the lens of sleep duration. But what about other dimensions of sleep health? After all, sleep regularity may be just as important as sleep duration in a number of contexts.

We once again teamed up with Ryan Rezai, a data scientist and student at the University of Waterloo, to visualize some MESA data. Once again, all the plots below were made by Ryan to highlight some intriguing trends in the MESA dataset. As always, we think there’s a lot of value in looking for pictures that can help you grapple with the complexity of multifaceted, complex phenomena like sleep.

Let’s start with the basics:

First we need to define sleep irregularity. We can do this in a number of ways. In the plots below, we’ve defined it in the ways MESA does—as either the standard deviation in total sleep duration (sd24hrsleep5) or the standard deviation in bed time (sdinbedtime5). 

So how does sleep irregularity, as defined above, relate to Epworth Sleepiness Scale self-reports (ESS) in the dataset? Like this: 

Looking at this, I feel pretty confident that there’s a trend here! As sleep irregularity increases, so does ESS, up until you get up to pretty profound variability in sleep regularity (180 minutes = 3 hrs standard deviation—you might be a shift worker at that point).

It’s even more remarkable when you look back at sleep duration and ESS (see last blog):

Not only does sleep regularity seem to have a clearer trend with ESS than sleep duration, it also seems to have a slightly higher amplitude effect, as shown on the y-axis. Right off the bat, this is a clue that we might be wanting to pay more attention to sleep regularity when we talk about the experience of sleepiness.

If we look at both bedtime irregularity and sleep duration simultaneously, we can notice something else interesting: 

Namely, that even for people sleeping quite a long time (e.g. 500 minutes), greater bedtime irregularity is linked to greater feelings of subjective sleepiness. 

What might be going on here?

We know a person’s subjective experience of sleepiness doesn’t always line up with how restricted their sleep has actually been. For instance, people on four hours of sleep a night tend to get worse and worse at reaction time tests, while their subjective sleepiness grows for a while but eventually levels off. Maybe irregularity makes people more reliably aware of just how sleepy and impaired they are because their irregularity means they’re more likely to be awake during periods of time when melatonin is at a high concentration in their body. Or maybe irregularity is having a dampening effect on their body’s circadian rhythms, making them more likely to feel exhausted and flat. There’s plenty of work to be done here in the future, but I’ll take off my Hat of Speculation now.

Beyond sleepiness

These 3-D sleep plots can be used to visualize more than just ESS. Take, for instance, this plot of total apneas over the course of the night as a function of sleep duration and sleep regularity: 

Ok, wow! That’s a clear picture, albeit perhaps not a surprising one in some ways. After all, you’d expect a longer duration of sleep to mean more opportunities for apneas. That said, it is interesting how, once you get above about 300 minutes of sleep (5 hours) or so, holding sleep duration fixed and increasing irregularity seems to correlate with increased apneas. 

Something similar appears to hold for sleep irregularity, sleep duration, and apneas per hour, with more sleep irregularity linked to a higher rate of apneas—at least when you ignore people sleeping around 200 minutes a night (which, to be clear, is probably not that many people):

Sleep irregularity and heart rate

Lastly, we might be interested in how sleep irregularity correlates with heart rate. After all, recent research has shown the risk of a cardiovascular event is more than twice as high in irregular sleepers as it is in regular sleepers. When we look at the irregularity in sleep duration, it sure looks like there might be something going on with average heart rate and irregularity in how long you sleep:

This trend also seems to hold when you look at the correlation between bedtime irregularity and average heart rate:

For both of these plots, the standard error is smallest between 0 and 180 minutes of standard deviation in sleep irregularity—and like we noted earlier, three hours standard deviation in sleep irregularity is a lot! (What’s going on when the standard deviation of bedtime irregularity is around 300 minutes? I sure as heck don’t know. But since that’s a standard deviation of five hours in sleep irregularity, odds are good that that’s not the typical sleeper.)

On the whole, it seems pretty clear: People who care about their overall sleep health shouldn’t sleep on sleep regularity.

With thanks to these resources:

Zhang GQ, Cui L, Mueller R, Tao S, Kim M, Rueschman M, Mariani S, Mobley D, Redline S. The National Sleep Research Resource: towards a sleep data commons. J Am Med Inform Assoc. 2018 Oct 1;25(10):1351-1358. doi: 10.1093/jamia/ocy064. PMID: 29860441; PMCID: PMC6188513.

Chen X, Wang R, Zee P, Lutsey PL, Javaheri S, Alcántara C, Jackson CL, Williams MA, Redline S. Racial/Ethnic Differences in Sleep Disturbances: The Multi-Ethnic Study of Atherosclerosis (MESA). Sleep. 2015 Jun 1;38(6):877-88. doi: 10.5665/sleep.4732. PMID: 25409106; PMCID: PMC4434554.

The Multi-Ethnic Study of Atherosclerosis (MESA) Sleep Ancillary study was funded by NIH-NHLBI Association of Sleep Disorders with Cardiovascular Health Across Ethnic Groups (RO1 HL098433). MESA is supported by NHLBI funded contracts HHSN268201500003I, N01-HC-95159, N01-HC-95160, N01-HC-95161, N01-HC-95162, N01-HC-95163, N01-HC-95164, N01-HC-95165, N01-HC-95166, N01-HC-95167, N01-HC-95168 and N01-HC-95169 from the National Heart, Lung, and Blood Institute, and by cooperative agreements UL1-TR-000040, UL1-TR-001079, and UL1-TR-001420 funded by NCATS. The National Sleep Research Resource was supported by the National Heart, Lung, and Blood Institute (R24 HL114473, 75N92019R002).

Circadian science Sleeping troubles Technology

One Simple* Rule for Understanding Your Sleep

My friend sent this to me the other day:

“If I go to bed at 10:30 pm, I often wake up at 4:30 or 5:30 am and can’t go back to sleep. But if I go to bed at 1:00 am, I fall asleep easily and can sleep until 9:00 am. Throughout the work week I’m running on 5-6 hours of sleep, and only on the weekend do I get to recover. Could it be that my natural cycle is just different from what social norms say?”

Reader, I pushed my glasses up my nose so fast and so hard that they are now deeply embedded into my face. A small price to pay for this gift of a question.

But before I can get into why I was so excited about my friend’s pain and suffering, I need to tell you about the two-process model of sleep.

The Two-Process Model, 10,000-ft-View Edition.

When you’re awake, you build up a “hunger” for sleep. This is your sleep homeostat, and it grows when you’re awake and drains when you’re asleep.

Two days of wake and sleep. Purple line goes up, you’re awake. Purple line goes down, you’re asleep. Shaded purple regions show sleep windows.

At the same time, you’ve got a circadian drive for alertness, which for day-adjusted folks tends to send its strongest signal for sleep in the middle of the night. The effect of both of these together is your sleep drive.

When the sleep drive hits an upper threshold, your body switches into “wants sleep” mode and tries to get you to go to bed. When it drops below a threshold, your body switches into “wants wake” mode and tries to get you to wake up. You can visualize this phenomenon as the zig-zag of the homeostat bopping between time-varying upper and lower thresholds set by the circadian clock. Hit the upper limit, fall asleep. Hit the lower limit, wake up.

Visualizing the two processes—homeostatic (purple) and circadian (orange/red).

Still with me? Ok, here’s the most important picture you’ll see in this whole blog post:


Well, there you have it! My friend’s sleep problems in a nutshell. No further explanation needed.

I showed this graph to another friend of my mine, and he sent back this:

Okay, so maybe a little more explanation needed.

Simple model. Complex phenomena.

Let’s talk about the two-process model in the real world. For one, people don’t instantly pass out when they hit the upper threshold for sleep. People can push sleep back in a lot of ways— staying out of bed, keeping the lights on, or, as was the case in one old sleep study, repeatedly dunking their heads in ice water. We can call this a “wake effort”—making an effort to stay awake in the face of the tyrannical rule of the two-process model. In fact, it might not really feel like that much effort if you’re having fun on the internet.

The blue-dashed line in the plot below shows what would happen if my friend stayed up a couple hours later than her body necessarily wanted to, moving her bedtime from roughly 10:00 pm (no wake effort) to a bit after midnight (with wake effort). The blue shaded region shows her sleep when she delays her bedtime in this way.

Purple line: falling asleep right when the two-process model says to. Blue line: Staying awake a little longer. Two different choices, two different sleep durations (blue vs. purple shaded area.)

Here’s the thing: That blue shaded sleep area, the one that starts a little after hour 24 (midnight) and goes until a little bit after 8:00 am? It’s wider than the shaded purple sleep area, which starts around 10:00 pm (hour 22) and goes until maybe 4:30 am. In other words, she’s sleeping more by staying up later—almost two hours longer.

You can eyeball it if you look where the lines are hitting the red waveform on the bottom. The purple and blue curves are chasing it as it’s going down, and the “stay up later” blue curve is hitting it at a significantly later point. That means more sleep overall (“good,” in theory), despite a later bedtime (“bad,” in theory).

What does this mean? Well, for starters, it hints at the enormous complexity you can start to get at when you mix two waveforms and thresholding conditions. The wiggly line of circadian sleep drive and the zig-zag line of homeostatic sleep drive, while deceptively simple, can interact to generate some wild phenomena.

Let’s explore a bit. I’m starting my friend’s sleep homeostat at a fairly high value, because she told me she tends to need an alarm to wake up during the work week (and feels pretty tired all of the time). If I started her at a lower sleep homeostat (waking up better rested), the difference between the two sleep durations becomes a lot smaller (purple: 7.7 hours of sleep, blue: 7.9 hours).

Starting my friend at a low homeostatic value. Now there’s really not a difference in sleep duration between her staying up a few extra hours or not—both the purple and blue shaded regions have the same duration.

Give her a higher starting homeostat, on the other hand, and the no-resistance-to-sleep-drive curve (purple) has her falling asleep around 7:30pm, waking up at midnight, and staying awake all night until she passes out again in the early morning, while the “wake effort” curve in blue mostly stays the same:

Now there’s a huge difference in sleep between the two options! They’re basically complements of each other.

These are wildly different scenarios, and they’re arising from me gently nudging a number up or down a little bit. Imagine what happens if we make those upper and lower thresholds—representing the circadian drive for alertness—act the way a real circadian rhythm does: shifting, stretching, and bending in response to the signals you give it during the day. Imagine weakening the signal from the body’s clock, or adding the effects of noise:

Circadian madness

Mathematicians who study this kind of stuff have done some pretty great work looking at just how much complexity can arise from the simple rules of a sleep homeostat and a circadian rhythm. In my own life, I think of the two process model most often when I wake up in the middle of the night. Ok, I tell myself. My circadian sleep drive is probably running a bit late. If I hang out in the dark a while, it’ll swing in and kick me back into sleep. And what’s great is that it reliably does.

“Are my natural cycles just different from social norms?”

But let’s get back to my friend’s question. If I had to make a guess as to what was going on with her based on the two-process model, I’d say her circadian clock is delayed relative to where she wants it to be (and that she’s often waking up in the morning with a high sleep homeostat due to the chronic sleep restriction from her work.)

That might seem to be a vote for her natural rhythms just not jiving with her work hours, and there’s probably some merit to this. She’s described herself as light sensitive, which may be a trait that predisposes you to being more of a night owl. That could be shifting her rhythms later, and making it so her circadian clock isn’t where it needs to be when her sleep homeostat drops to a low value.

Yet one of the great things about circadian rhythms is that they can change. If they didn’t, we’d never get over jet lag. We’d stay on our starting time zone schedule for all time, regardless of where we went, when we slept, or when we went outside. The fact that our clocks can be disrupted also means they can be fixed.

Another way of thinking about it: My friend’s weekend rhythms would make her a pretty darn early bird if we abruptly transplanted her three time zones west to California. What’s the difference between living in California versus her home city on the East Coast? Well, a lot of things, but the most important one for her circadian clock is the difference in her light exposure.

So I’ve got her trying out a new circadian regimen to help herself sleep during the work week. My hope is that the two-process model, along with the fancier flavors of it that have spun up in research and at our company, can help people in the real world understand their sleep better. Simple rules can interact to make complicated behaviors happen, but at the end of the day they’re still simple rules. And simple rules might just have straightforward solutions, too.

Circadian science Technology

Scientific Tests

Algorithms are really easy to mess up. Take your pick for how: overfitting to training data, having bad training data, having too little training data, encoding human bias from your training data in the model and calling it “objective”. Feeding in new data that’s in the wrong format. Typos, subtle typos, nightmarishly subtle typos. Your cat stepping on the keyboard when you’re out of the room.

Having done this for a while, my first impulse any time I get amazing performance with an algorithm is to be deeply suspicious. This isn’t because algorithms can’t be incredibly powerful; you really can get amazing performance if you do them right. But you can also get seemingly amazing performance if you do them wrong, and there are a lot of ways to be wrong.

The core issue, I think, is that there are so many choices involved in the making and maintaining of an algorithm, and if the algorithm is trying to do something complicated, those choices can have complicated downstream effects. You can’t readily anticipate what these effects might be (this problem is hard enough that you’re building an algorithm for it, after all), but your brain tells you that there can’t be that much of a difference between a threshold being 0.3 and that same threshold being set to 0.4. So you blithely make the change to 0.4, expecting minimal effects, and then the whole thing just collapses underneath you.

I’m saying this because, while developers on the whole have gotten pretty on board with the concept of unit tests and test coverage for code, I’m not as sure about what currently exists around tests for data science algorithms in biology, medicine, and health. I’m not talking about tests that confirm a function gives the number we expect it to based on an old run of it (e.g. asserting that f(3.4) = 6.83 because we ran it once with 3.4 and got 6.83 as the answer). Those kinds of tests basically act like a flag that something’s changed, and if you change your function f, you can just paste in the new output from f to make the test pass.

I’m talking about using a lot of data — a representative sample of what you’ve collected— in your code’s tests, to assert that some macro-property of the algorithm’s output is preserved. If you make a change to the algorithm, and that macro-property changes, your code should let you know about it. If this sounds like one flavor of functional testing, that’s because it is. But the key thing I’m arguing for is that the performance of data science-components of a product— code two steps removed from sklearn or torch or R—be tested functionally as well.

Let’s talk through an example. Imagine you’ve got an algorithm for distinguishing sleep from wake over the course of the night. In my own experience, algorithms of this kind can have an unfortunate tendency to flip quickly, in response to a threshold being changed, from thinking a person slept a lot over the night, to thinking that person barely slept at all. There are ways to address this, and it’s not always an issue, but it’s still something to look out for.

So how could you look out for it? Take a representative sample of data from a group of sleeping people, and add a test that runs your sleep algorithm on all of them and asserts that everyone gets detected to have at least some baseline amount of sleep. This way you don’t have to worry that changes to the algorithm made some people better but other people unexpectedly, dramatically worse. You can rest easy because assurances of this kind are built into your development pipeline.

At Arcascope, a lot of our scientific tests center around the mean absolute error of our prediction of melatonin onset. We want to predict when somebody’s melatonin onset is happening so we can map it to other quantities of interest: minimum core body temperature, peak athletic performance, peak fatigue, you name it. But we also want to do continual development work on these algorithms, to keep bringing the mean absolute error down over time. How can we make sure that our changes actually make things better? How can we make sure that a change that improves performance for some people isn’t sabotaging others? Tests that confirm the properties we care about are preserved, no matter what we do in the backend.

I mentioned above that I don’t know what’s out there elsewhere in digital health, and I don’t. Maybe a lot of people are writing tests of this kind in their code! All I know is that when we realized we could do this— add fundamental performance checks for the algorithms that make up our backend, embedding deidentified human data from our studies directly into our testing suite— it was a very cool moment. It helps me sleep a lot better at night knowing those tests are there.

Which is good, because sleeping more means I’m introducing fewer nightmarishly subtle typos to our code. Wins all around.

Circadian science Technology

Biophysics for Better Living 2.0

In my last blog post, I talked about the power and potential for biophysics engines to contribute to clinical care and ragged on my own ability to play video games. The short version of it, if you don’t have time to circle back: If you’re going to make a digital twin of somebody, base it on a system of differential equations that captures how the physics of the human body work. Also, I’m terrible at video games.

I talked in that blog post about how you can use a biophysics engine—a.k.a. known properties of how the body works, codified into equations— to estimate things that are hard to measure because they’re hard to reach. Things like the firing rates in your ventral and dorsal suprachiasmatic nucleus (SCN), which are tucked pretty deeply inside your brain. You’re not going to be able to easily peek and see what the SCN is up to (at least not right now, in 2022), but you can figure out what your SCN is likely up to passing the same inputs your real SCN got into a model of the SCN.

Woke up really early and got a blast of light? That probably sped up your clock. Stayed up late and kept the lights on overhead? That probably slowed your clock down. Where did your clock wind up at the end of it all? That’s a question that hinges on the details of when, how long, and how bright your light exposures were. In other words, it’s a question for the mathematical model.

But there are other questions you can ask a model, beyond “where am I now?” You can ask “what if?” What if I were to stay up the entire night in bright light? What if I dimmed the lights at 3:15 pm? How would these changes affect the state of my internal clock? And how would that changed state affect downstream outputs, like the timing of my peak fatigue or my peak performance?

When we talk about what we do at Arcascope, we say we do circadian tracking and circadian recommendations. The tracking is using a biophysics model to answer the question “where is your circadian clock right now?” The recommendations come from repeatedly asking the model “what would happen if you did (insert series of behaviors here)?” Asking a lot of questions gets you a lot of answers, which you can then pick from based on which ones best meet your goals, whether that’s sleeping more, adjusting faster, or being maximally alert at a specific time.

It’s all still biophysics models, but in one use-case, we’re putting them to work to capture reality, while in the other, we’re trying to capture a whole swath of possible futures. A list of potential realities to choose from as you scope out the rest of your week. Better living by picking the best answers to “what if?”— I think that’s pretty neat. And it keeps me up at night thinking—what if we could do this for other systems, besides sleep and wake, too?

Interested in beta testing our app? Send us an email!

Circadian science Technology

Biophysics for Better Living

One of the highest stakes moments of my life was the time in graduate school when I was (badly) playing Donkey Kong at Pinball Pete’s Arcade, and a very, very good Donkey Kong player came over to watch me play in silence. I’d just spent six dollars worth in quarters dying repeatedly on the first floor of the first level, but something about his judging, appraising eyes summoned the fireball-hopper within.

I made it past the first level, then the second, then the third—an absolute record for me—and I had almost cleared the fourth when the pressure finally got the best of me. Still, it felt like a tremendous achievement. I like to think the good player and I shared a nod of mutual respect as he elbowed me out of the way and proceeded to spend the next forty-five minutes on an uninterrupted victory run. 

Besides letting everyone know that I need to get out more, this story is an opportunity to talk about a kind of math that should be more integrated into clinical care, and isn’t— yet. There was a model of physics deciding what happened in that game, a system of differential equations for converting my button mashing into movement across a screen. And just like the inputs I gave the system could be transformed into outputs that captured the physics of motion, so too can the inputs we give our bodies be transformed by a realistic biophysics model to give us outputs that capture the ways our bodies work without the need for invasive tests. 

Let’s back up: At a very high level, what’s going on in a video game is that you’ve got prescribed rules for how to update all the positions of all the moving pieces from one frame to the next. In the next blink of an eye, where should Mario go? If I’m pushing the joystick to the left in that instant, nudge him a little to the left. If I’ve just hit the jump button, nudge him a little bit upwards. If I’m mid-jump, make the force of gravity and the speed he took off the ground fight it out to see if he should be nudged a little up or a little down. 

The whole set of rules for making these decisions on how to nudge can be called a physics engine. Passing the history of my joystick spins and button presses into this physics engine gets you an output that describes how Mario moves, how he jumps, and how far I make it in the game before an untimely demise. This type of math is all over the place in games, CGI, and user interaction elements (like views that scroll and bounce). But there’s no reason a physics engine has to be restricted to visuals, or video games, or the physics of movement. You can have a physics engine that captures ion channels open and close, for instance, or how neurons talk to each other in the brain. 

Which brings us to your body’s internal clock. Circadian rhythms are a massive pain to measure experimentally; it’s one of their defining properties. It’s hard to know what’s going on in the brain without looking at it, and it’s hard to get a glimpse of it without getting pretty invasive. But if you know the rules for how the clock part of your brain works— “the differential equations describing the suprachiasmatic nucleus”— you can pass through a history of inputs to the system, and get out an output that describes how the neurons are communicating with each other at a specific point in time. Pass in what my lighting history has looked like over the last four weeks, and I can tell you what time my brain would cue melatonin to start rising in my body tonight. 

So here’s why I think we should be using differential equations in medicine more: they can help you see things you otherwise wouldn’t be able to see. You can use a system of differential equations to track how the brain would react to the inputs you give it, even if you can’t readily experimentally measure brain state, just how a person with access to Donkey Kong’s code could track how far I got on my amazing run, even if someone blocked off the console screen with a tarp. Knowing the rules means you can measure things in the dark. 

And there’s another reason, too—personalization. By tweaking the rules of your video game physics engine, you can change the way the gameplay works. Cut gravity in half, and suddenly Mario’s soaring with every jump. Double it, and he’s a sitting duck for everything Donkey Kong throws at him. As the importance of individual differences in light sensitivity and fatigue management become better understood, biophysics engines are ready-made for customization. Want to better fit the data of someone who’s hyper-sensitive to light? Crank up the light sensitivity parameters. Got genetic data on them? Plug that into a molecular model of their rhythms

Biophysics engines can be the key to a true digital twin, helping us look inside a simulation of ourselves in ways we can’t look inside our real selves. There’s potential for these quantitative methods in medicine beyond sleep and circadian rhythms, too. Who knows? Maybe someday a model of this kind could be useful for understanding how the (fun) stress of my epic Donkey Kong run affected my mood and memory. Or at the very least, how my body processed the enormous bubble tea I drank in triumph right after. 

Want to learn why we don’t think the answer to circadian estimation in the real world is to just throw an artificial neural net at it? Check out this blog post by CTO Kevin Hannay. 

Circadian science Interviews Sleeping troubles Technology

Interview with Dr. Cathy Goldstein

We sat down to talk with Dr. Goldstein on what she sees as the future of wearables in the sleep clinic. Enjoy!

Thank you for taking the time to meet with me today! Could you take a minute to introduce yourself and say a little about what your job is?

Cathy Goldstein, MD. I’m a neurologist, Associate Professor of Neurology here at Michigan Medicine, and my clinical practice is entirely in sleep medicine.

You’ve done work with [Arcascope CEO] Olivia on wearables, and you’re heavily involved in the AASM and consumer sleep technologies. Where do things stand today for consumer sleep technology and wearables, and where do they need to go?

These devices really started becoming prominent about 8 years ago. I remember I was right out of my fellowship at that time, and I thought, wow, these will be so awesome for when a patient sees me in the clinic: I have all these questions about their sleep and sleep timing, and how they sleep on the weekdays vs weekends vs vacation, and they’re in the position of having to try and remember what their problems have been over the long time they’ve been waiting to see me.

And the problem with a sleep log [that someone fills out manually] is that it’s hard to keep up with every day. In this busy world, you need more of a passive degree of tracking, particularly over time. So when wearables came out I was like, wow, you know, wouldn’t it be cool if when the patient shows up I can just take a look at their data in the last six months and see their sleep patterns.

The problem as far as adoption has gone, is that we don’t know how accurate they are at tracking sleep. So, there’s a medical grade version of consumer trackers called “actigraphy”, although they’re not completely accurate, we KNOW how accurate they are. And so, particularly in science and medicine, what is known is better and less scary—even if the performance isn’t great— than the unknown.

There are a lot of unknowns in regard to consumer-targeted sleep tracking because there’s not a lot of peer-reviewed literature about the performance of these. It’s growing, but even when the accuracy does get reported in a paper, oftentimes the algorithms change with updates, or the hardware changes a little bit with each iteration. The big thing is—we just don’t know. And that’s really prevented adoption in research and clinical practice, unfortunately.

What are the biggest misconceptions about sleep wearables that you see out in the world?

The biggest misconception is that they’re totally inaccurate, that they’re inferior to actigraphy for some reason, because the data that we have so far doesn’t necessarily suggest that. For many of these devices, we just don’t know, but the ones that we do know about appear to perform similarly.

I’d say the other misconception is people say, “My watch tracks my sleep.” Your watch doesn’t track sleep: your watch tracks your heart rate or your pulse, and your watch tracks movement. And then, math estimates your sleep from that device. So, that’s a big misconception of people telling me in the clinic, “My device is telling me I don’t have REM, so my sleep is really bad.” Sleep stages like REM, NREM are polysomnography EEG constructs. So, to use a consumer wearable sleep tracker’s components of dream sleep, light sleep, deep sleep, interchangeably with PSG-defined states— we’re just not there yet in my opinion.

I do think the estimates are getting better, and they do use properties of heart rate variability that we know change in the different sleep stages, but still, those [stages] are defined by brainwaves and eye movements. So, it’s very hard to recapitulate that. The question is, “Do we even need to track that on a daily basis?” What we’re really looking for, and what the clinical and research world wants, is to track on a longitudinal basis— objectively— just sleep-wake patterns day to day.

How do sleep trackers fit in with sleep clinics right now?

People sometimes think that doctors just don’t want to look at sleep wearable data because we don’t like it—that’s not true. A lot of us love seeing this data and love seeing sleep patterns over time, particularly if a patient can tell me “Hey you know, you treated my sleep apnea with this CPAP machine, and look at the change in my wearable tracked sleep.” One of the massive problems, though, is that clinical practices are really, really busy. We have a lot of patients that need help, and we need to give everybody high-quality care.

This means we have a high volume of work and most of our work takes place through something called the Electronic Health Record, and at this point, there is no real way to interface the data that comes from consumer wearable devices with that Electronic Health Record. So, we really don’t have a good way of integrating any of this into our clinical practice. It’s not that we are absolutely opposed to using this as an adjunct, particularly the ones that have some idea about performance, but we don’t really have a way of getting it in there to make it an easy part of our workflow. Patients will send me screenshots, or they will show me their app in the clinic.

I think medicine moves slower than consumer-geared technologies for a multitude of reasons, including security issues, but I do think that one day we will get there. And again, we’re not using these as diagnostic tools; we’re using them as an adjunct to our clinical decision-making. So, as long as they are reasonably accurate, and we overall know how they work, I think a lot of doctors would like to adapt them into their practice as long as we have a way of seamlessly integrating that data into our workflow.

Do you think we’re going to see circadian rhythms enter the clinic more in the near future?

Yeah, and what I would hope is that they don’t just enter the sleep clinic, but they enter the wellness and general health area as well. I think a lot of us are doing all of our body systems a big disservice by living in desynchrony with both our central clock and peripheral clocks.

When my friends and family ask, “What are some of the best things I can do for my sleep and circadian health?” I say “Wake up at the same time every single day.” That’s going to entrain your circadian phase. And hopefully, you’re eating in line with that (and you’re not eating when you’re supposed to be asleep), and when you’re getting light when you wake up. We are all undergoing mild degrees of circadian disruption by varying our wake-up time and getting as much light at night as we are.

Definitely. Especially with screens. I was one of the people who thought my phone screen couldn’t possibly affect my quality of sleep.

Exactly! And I think people are becoming more aware. I mean when you talk about intermittent fasting, that’s kind of a chronotherapeutic measure. It’s so simple, but people get so excited about it. It’s like yeah, don’t eat when your body is biologically prepared to be asleep.

You’re right that it’s incredibly simple, and I think people really respond to small changes that make a big effect on their health and wellness.

And it’s not magical, it’s timing. It’s literally all about timing.

What do you believe the future holds for sleep technology? What are you most excited about?

I’m just hopeful for a day where we change the way clinical evaluation works now, where a patient might be waiting for me to see them for months and months, and then I see them, give them instructions, send them home with sleep logs so I can see how they’re doing, and tell them to come back to the clinic in 6 months because I’m booked out I can’t see them sooner.

I want to get to a point where I see them, and at that initial point of care I know what their last 6 months of sleep looked like, and then I can come up with an intervention that’s precise for them and that’s also adaptive based on how their sleep looks in response to intervention.

Then possibly, when they do go home, we can change that intervention, maybe in an automated way, maybe with me being able to interact with an app, whatever it may be – but instead of writing down instructions and giving people medications, we’re using the mobile application as a prescription to really make patient-specific interventions that are based on wearable data.

So, it’s important to make it less labor-intensive for the patient because then it’s less likely to be done correctly or be done at all.

Exactly, we’re talking about behavior change here. That’s one of the cornerstones of so many diseases we take care of in medicine: they’re due to things that with behavior change could be different, and behavior change is hard. One of the behaviors that we’re really, really good at in current society is working with our devices and our apps. So, I think it’s just a no-brainer as far as delivery goes, I think it’s really time to make this stuff an adjunct and a helper in healthcare.

What have you seen in the clinic in the age of COVID?

I’ve seen a dichotomy. I mean I don’t think we’ve ever collectively (people my age, middle age, most of my patients) have gone through a stressor like this. So, there are quite a few people who had significant insomnia; there are people that had COVID that had major health disruptions during and following that, including fatigue during the day. Then there are people who actually had marked improvements in their sleep because they can sleep according to their clock, and they have more time due to less commute and so they can extend their sleep a bit.

There have actually been patients who go off alertness-promoting medications, and are a bit happier with their sleep. There was a great article about a gentleman who had always felt confined to the service industry because he was a night owl. He was always kind of stuck, like a bartender/server, and during COVID (because hours were more flexible with work from home), he was able to go into other industries and have regular hours at the time he selected, and have a more stable work path. Which is what he wanted. You know, we shouldn’t chronotype shame people. Just because you’re biologically late you shouldn’t be at a disadvantage in life, and that’s helped a lot of people’s sleep.

It’s nice to hear some good news come out of the year 2020. That’s all the questions I had lined up, but do you have anything you would like to add?

I think there’s also, what can wearables NOT do. So, we don’t know how accurate the oximetry is yet, we don’t use this as a way to diagnose sleep apnea, I don’t think that sleep staging is something we should rely on. It’s that sleep disordered breathing, is not really something we have not been able to pick up yet, so when people have some episodes of sleep snoring or gasping, feeling sleepy during the day – even if your wearable says you sleep great, that is a limitation and you should still see your doctor.

So, if you feel something isn’t right, it’s best to go in and talk to your doctor.

Exactly, trust your body. The manifestation of diseases is how you feel, not just how your numbers are.

Circadian science Technology

Yesterday’s Weather and Your Health Data

I love my Apple Watch.

The ability to track my exercise, heart rate, activity levels, and sleep has enabled a real awareness of how my physical and mental health changes over time. The ability to track personal health data over long time periods outside of laboratories is one of the most exciting developments of the last decade. I believe this data will usher in a new era of personalized, precision health which just wasn’t possible in the past. At Arcascope, we are at the forefront of developing algorithms to turn the data collected by wearable devices into insights that improve people’s lives. 

With that being said, the current state of things just isn’t all that satisfying when you think about what’s being left on the table. So much of the data being collected is uninterpretable. Knowing my current heart rate is cool, but what can I do with that information? The part of this that bothers me the most is that so much of this data is focused on the past. 

Here is a screenshot from Apple Health showing my sleep over the last month. You can see that I had some wake periods at 3 am at the beginning of the month. But how does this information really help me? 

Sleep tracking in particular reminds me of a weather app that can only tell you yesterday’s weather. Clearly, a weather prediction service that could only tell you the weather from 24 hours ago wouldn’t do well against the Doppler radar. It is useful to be able to say exactly how hot it was yesterday, and interesting to know how that compares to years past, but I really want to know if I should bring an umbrella with me when I leave the house. 

I can tell that I didn’t sleep well last night from the fact that I am feeling tired. Having a device to quantify exactly how poorly I slept can be useful for tracking long-term trends, but it isn’t all that useful on a day-to-day basis. 

Another snapshot from my Apple Health data. Doesn’t this remind you of a weather app pointing out how this weather’s month compares to historical trends? What about today? Or how about tomorrow?

Okay, enough of the weather prediction analogy. I’ve already pushed that analogy further than I should. First, unlike the weather, we actually have control over our behavior, and what we are doing now will change the forecast for our physiology tomorrow. Also, these variables are much more predictable than weather. 

The technology we have developed at Arcascope can answer questions like: 

  • What separates the days where I am at my best, from the ones where I am struggling? 
  • How can I alter my behavior now so that I will sleep better tonight?
  • When is it best for me to stop drinking coffee for the day? 
  • When will it be best for me to study, exercise, eat and relax? 
  • When is the best time to take my medication to minimize side effects? 
  • When should I avoid high stakes activities because my chances of making a mistake are highest? 

We believe this is the future of personalized health tracking. We also think it’s a heck of a lot more exciting than looking back at yesterday. 


Circadian Phase Estimation and Deep Learning

One of the most common questions we get at Arcascope is…

“Can’t you just do circadian phase estimation using machine learning?”

Living in the data age, we have become used to thinking that big data and machine learning can do just about anything. In this post, I will break down some of the unique challenges for circadian phase estimation with an eye towards machine learning techniques. I’ll also do a brief review of the previous attempts to apply machine learning to this task.


Wearable Headaches (and how to fix them)

So you want to get somebody’s internal time from a wearable…

Let’s talk about wearable data. On the one hand, wearables are an incredible innovation, allowing self-quantification and anomaly detection with unprecedented ease, at unprecedented scales. 

On the other hand, they’re a data science nightmare. Or three nightmares, really. 

Nightmare #1: All the devices are different, and you have to use different ways to get raw data off them. 

Sure, apps like Apple Health that act as clearinghouses make this easier for you. But you can’t use Apple Health for everything. Sometimes, wearables require permission to be granted for you to access their full data. Sometimes, wearable companies go out of business after you’ve built an infrastructure to work with them. 

Can you process heart rate signals from two wearables using the same algorithm? What if they decide what counts as a “step” in different ways? What if the firmware changes? People have certainly thought about these questions, and that’s the whole point: you have to think about them. The effort of keeping track of everything adds up.