Joint ESRS-SRS Seminar Spring Edition – Circadian Biology: Converging Pathways from Bench to Bedside – 20 March, Q&A Highlights (Part 1)

PZ: The most practical way would be with a device that measures activity, heart rate and temperature and light exposure. So any combination of these -activity and light being essential can provide measurements of sleep (timing, duration, fragmentation and regularity), as well as circadian rest-activity measures and if add heart rate, temperature etc-can get 24 rhythms of these as well.

MM: do you have a concept for how to define sleep health and circadian ‘measures’? I do not thus we first have to define these concepts before we can speculate on their assessment. I can imagine that by ‘circadian measures’, maybe you mean entrained phase or amount of social jetlag (thus inferring more or less internal synchrony accordingly).

PZ: There are several important projects in the pipeline to address your excellent question. Much of the health data is collected in the EHR systems, including 24- hour data (blood pressure, heart rate etc) in hospitals. Challenge is to prospectively organize the various sources in real time, so that its easier to extract and analyze. One of my junior colleagues is doing that in primary care-to use the electronic health records to assess clinical outcomes from sleep disorders or sleep/circadian measures. First these questionnaires/measures need to be included in the EHR.

MM: I guess that if you hypothesise that most people’s chronotypes are within a few hours of each other, then you can apply the methods used by among others Chen, … and McClung (https://doi.org/10.1073/pnas.150824911) broadly. With a large enough N, you can see markers that are rhythmic with a high enough amplitude. All of these devices/tests will have time stamps, thus it will give a nice time series for analysis.

PZ: Absolutely. I think modelling that allows us to look at phase relationships (between external with internal, but also among internal rhythm at different times of the day) will be more informative that phase of a single rhythm.

MM: That would hypothesise non-sinusoidal or unusual shaped oscillations which is likely but regarding which we have little experience. I recall that Päl Westermark has been working on this problem (how to evaluate and analyse, e.g. sawtooth oscillations). See here: doi: 10.1177/0748730414553029. Further, he has developed other methods. How to think of the basis for such an idea? When you look at phase plots for transcriptomes, you often see a dominant phase where the majority of genes are phased within 3-6 h of each other. But then there can be a second peak, and if that is regulated by a different clock-regulated output pathway, perhaps even a different zeitgeber, then it is possible to imagine that one set is on one phase while the other wavers.

PZ: It will depend on what you intend to use this data for-as far as accuracy, scalability, light accuracy etc. There are research grade actigraphy/light sensors, as well as consumer (like Oura-which has been used mainly for sleep-but also can infer sleep wake and cardiovascular rhythms. For people with diabetes -there are ambulatory glucose sensors and for cardiovascular disorders, heart rate and blood pressure. All of these can be used outside of the clinic for many days.

PZ: Biomarkers that are scalable to detect phase and amplitude and predict clinical outcomes of circadian-based treatments.

PZ: The circadian rhythm sleep-wake disorders are based on the chronobiological differences. Other sleep disorders, like insomnia, restless legs etc could benefit from including circadian/ chronotype information.

PZ: Its happening and perhaps the key- as wearable and nearable devices are popping up and cheaper- the key is to use this large amount of continuous data to assess circadian rhythm measures with health outcomes-like metabolic health (glucose monitoring with heart rate, actigraphy).

PZ: I hope in 5 years. We are conducting studies using pupillometry and when combined with actigraphy/light algorithms to infer DLMO-we can use these as diagnostic tools.

MM: This is of course, one of the most complicated problems, this here that you bring up. It has not only scientific but also social, economic and political sides to it. Scientifically, an ideal situation would be to take an individual, objectively determined chronotype and schedule school and work accordingly. But in practical terms, everything is wrong with this. Parents who have to get to work. Teachers whose chronotypes are earlier than all those adolescent students. How to pay for multiple parallel school groups based on a distribution of chronotypes. How to facilitate optimal communication between work colleagues if the actual work times only overlap briefly due to this scheme. What if some courses at a University (where school times seem somewhat more flexible) were organised according to the chronotype of the Professor and student combined? It would lead to a generation of students trained in something that might be less than optimal based on the avoidance of social jetlag. Honestly, I am not sure how to think about this and how to synthesise the various aspects.

MM: As a basic scientist, I’m going to stick with the concept of circadian organisation: I think that we have the opportunity to really start to understand how circadian systems are put together. When we can define internal synchrony, then we can define desynchrony. Or maybe it goes the other way around. I think that will be an important insight and will lead to many opportunities. But for now, I would also note as an opportunity the influx of specialists from other fields into our world of chronobiology. Take, for instance, the translational paper by Eng Lo’s group (doi: 10.1038/s41586-020-2348-z), figuring out important aspects of stroke pathology and time of day with big implications for treatment. This has led to a massive effort to delve into stroke and circadian clock in a big way and it will soon be implemented into treatment of stroke. This will happen with many diseases in the coming years because, as we know, the clock controls almost everything.