Assessment of Circadian Rhythms

Introduction

The focus of this chapter is to review several commonly used methods for the assessment of circadian rhythms as it relates to circadian rhythm sleep-wake disorders (CRSWD)¹. It will also touch on recent research on the development of new biomarkers of the circadian system in humans. It is not the goal of this chapter to provide specific details of the methods to assess circadian rhythms, rather it is to review the key factors to consider when selecting a method and when collecting and interpreting the data from these measures. Reference to works that do provide more detail on the methods themselves will be provided within each section.

To set the stage, the reader must first have an understanding of the circadian system, the factors that influence (or mask) the outputs (or markers) of the circadian system, and the terminology used to describe the circadian system. There are numerous output rhythms of the circadian system that can be assessed to determine the phase, amplitude and period of the circadian system. A graphic representation of these circadian concepts is provided in Figure 1, the difference in the level between peak and trough values is the amplitude of the rhythm. The timing of a reference point in the cycle (e.g., the peak) relative to a fixed event (e.g., beginning of the night phase) is the phase. The time interval between phase reference points (e.g., two peaks) is called the period ².

Figure 1. Parameters of Circadian Rhythms.

A representative circadian rhythm is depicted in which the level of a particular measure (e.g., blood hormone levels and activity levels) varies according to time. The difference in the level between peak and trough values is the amplitude of the rhythm. The timing of a reference point in the cycle (e.g., the peak) relative to a fixed event (e.g., beginning of the night phase) is the phase. The time interval between phase reference points (e.g., two peaks) is called the period. The rhythm shown persists even in continuous darkness (i.e., is free running). ( (From Vitaterna MH, Takahashi JS, Turek FW. Overview of circadian rhythms. Alcohol Res Health 2001;25:85-93; with permission)

Assessment of circadian rhythms in clinical practice can serve several functions, the most important of these is the ability to determine the extent of circadian disruption or the misalignment of circadian rhythms with the external environment. Another practical reason is to precisely time circadian based treatments such as bright light exposure and exogenous melatonin administration. The timing of these treatments relative to the internal circadian clock will determine whether they result in phase advances or phase delays and this response can be expressed graphically as a phase response curve ^3-7. A simplified example is that light exposure prior to the core body temperature minimum results in phase delays of the circadian clock and light after the core temperature minimum results in phase advances⁷. Mistiming of circadian based treatments like light and melatonin may worsen circadian disturbance and exacerbate other symptoms. Currently the most common way that clinicians determine the appropriate time to administer these treatments is by setting a time relative to either sleep start or end. This strategy makes certain assumptions about the phase relationship between various circadian rhythm outputs such as sleep, melatonin onset and core body temperature minimum. Another reason to assess circadian rhythms is to better understand the pathophysiology of circadian rhythm disorders. There is recent evidence to suggest that not all individuals that meet the current clinical criteria for delayed sleep-wake phase disorder have a delayed circadian phase of melatonin, even though their sleep-wake cycle is delayed ^{8, 9}. Since many circadian based treatments are currently scheduled relative to sleep timing, the lack of a delayed melatonin rhythm in some patients with delayed sleep-wake phase disorder could make appropriate timing of circadian based treatments difficult and could explain why there have been mixed results in the literature in studies using circadian based treatment is this population¹⁰.

Finally, the need to assess the timing, amplitude and alignment of the various outputs of the circadian system is supported by growing evidence, to suggest that misalignment of circadian rhythms in central and peripheral tissues can not only lead to poor sleep but also poor physical and mental health outcomes ¹¹.

Circadian Rhythms

The sleep-wake cycle is perhaps the most prominent circadian rhythm in humans. Circadian rhythms, which are present in virtually all physiological and behavioral functions, and are generated by a circadian pacemaker located in the suprachiasmatic nuclei (SCN) of the anterior hypothalamus ^12-15. The circadian timing system is conceptualized as three distinct components: a circadian oscillator with a rhythm approximating 24 hours, input pathways for light and other stimuli that synchronize the pacemaker to the environmental light/dark cycle (zeitgebers), and output rhythms which are regulated by the pacemaker. Circadian rhythms are intrinsic to the organism and in the absence of synchronizing stimuli, they will continue with a period of approximately 24 hours ¹⁶. Circadian rhythms are synchronized to the environmental light/dark and social/activity cycles by daily adjustments in the timing (the phase) of the rhythm¹⁶ this process is called entrainment.

In humans, light is the most effective synchronizing agent for the clock. Photic information reaches the circadian system via a direct pathway, the retinohypothalamic tract ¹⁷, and an indirect pathway via the intergeniculate leaflet ¹⁸. It is now recognized that the primary circadian photoreceptors are the intrinsically photosensitive melanopsin containing retinal ganglion cells ^{19, 20}. Furthermore, both light-induced phase shifts and melatonin suppression are most sensitive to short wavelength light at approximately 460 nm^21,22. In addition to light exposure, exogenous melatonin administration and physical activity have also been shown to induce changes in the phase of circadian rhythms ^{23, 24} .

Regulation of the human sleep-wake cycle is generated by a complex interaction of endogenous circadian (process C) and homeostatic (process S) processes as well as environmental factors ^{25, 26}. Process C is driven by the circadian clock that promotes wakefulness during the day and facilitate the consolidation of sleep during the night ^27-31. Maximum alertness occurs in the early evening hours when the drive for sleep is also highest ^32-36. The amount and depth of sleep is instead determined by the sleep homeostatic process S. The homeostatic process of sleep accumulates as a function of prior wakefulness; currently sleep is the only known way to reduce this homeostatic drive after extended wakefulness ^{25, 37, 38}. Therefore proper alignment between circadian timing of sleep-wake and other behaviors such as feeding are essential for optimal sleep, waking function and health^11,39-45.

Common Measures Used to Assess Circadian Rhythms

Circadian rhythms can be observed in almost all physiological functions, but in humans there have been three main outputs used to assess circadian rhythms to date; melatonin levels, core body temperature, and rest-activity cycles. Given limited availability of some of these objective measures in the clinical setting, other tools such as sleep diaries and chronotype questionnaires can be used to aid clinical diagnosis of CRSWD’s.

In the clinical setting, most markers of circadian timing are collected in the field/home while patients are going about their usual daily activities. However, given the masking effects of light, activity (including sleep-wake), and feeding on some of these measures, the use of specialized protocols within the laboratory setting specifically designed to minimize the influence of these factors is often preferred to obtain an accurate assessment of the underlying circadian timing (i.e. constant routine, forced desycnhrony ^{33, 46-48}). The down side to a laboratory-based approach is that people do not “live” in a laboratory, and understanding the interaction between circadian, environmental and behavioral factors may be important for diagnosis and management in clinical care. Given this, some key factors to consider when instructing patients on the collection of the measure and in the interpretation of the data are provided for each circadian marker discussed.

While the measurement of the dim light melatonin onset (DLMO) is encouraged in the diagnostic criteria for some CRSWD’s¹, there are no specific guidelines for which technique to determine DLMO is best for clinical purposes, and there are benefits and disadvantages to each method. In general terms, when selecting a method to determine circadian rhythms, consideration should be given to what you are trying to measure, what the population is, whether the samples will be collected at home or in the laboratory, the level of invasiveness, the sampling period and frequency to be assessed whether it be a single evening, a full 24 hour period or even assessment over multiple days. Cost is also a major consideration, many of these tests cannot currently be claimed with medical insurance and are an out-of-pocket cost, which in some cases limits these tests to those that can afford it.

Sleep Log/Diary

Keeping a daily sleep log/diary for 14 days (minimum 7 days) on both work/school-days and free-days is part of the Internal Classification of Sleep Disorders (ICSD) diagnostic criteria for the following CRSWD’s: delayed sleep-wake phase disorder, advance sleep-wake phase disorder, non-24 sleep-wake rhythm disorder (minimum 14 days), irregular sleep-wake rhythm disorder, and shift work disorder ¹. Sleep logs allow the clinician to characterize the pattern of sleep-wake. Sleep logs should include information about the day of the week, whether it is a work/school or free-day, when the person went to bed, when they fell asleep, when they woke up, and when they got out of bed. It is also useful if the sleep log contains information about other behaviors such as when the patient took medications, consumed caffeine or alcohol, when they napped or exercised. The American Academy of Sleep Medicine has a sample two-week sleep diary available on the web⁴⁹. Another useful piece of information that can be captured on the sleep log is the start and end of electronics use each day. Electronic devices emit light, and depending on the activities they are being used for they may be alerting (video games, movies), which could exacerbate sleep-wake and circadian disturbance in those with sleep disorders.

Circadian Chronotype

A tool frequently used as a proxy of the timing of the circadian system is circadian preference or chronotype assessed via questionnaires^{50, 51}. Circadian chronotype is typically defined by a person’s preferred time to conduct daily activities⁵⁰, and or the timing of sleep^{51, 52}. The most commonly used questionnaires are the Morning-Eveningness questionnaire⁵⁰, the Munich Chronotype Questionnaire (MCTQ) and the Munich Chronotype Questionnaire for shiftworkers (MCTQShift) ^{51, 53}. While these questionnaires align reasonably well in healthy controls with other measures of the circadian system such a CBT and melatonin levels^54-58, in patient populations these relationships are less clear.

These questionnaires do not measure circadian rhythms per se, but they can still be useful tools to identify extreme early or late types and to characterize concepts such as ‘social jetlag’⁵⁹. The International Classification of Sleep Disorders criteria now suggest that standardized chronotype questionnaires can be useful tools to assess chronotype in those with advanced or delayed sleep-wake phase disorder ¹. So even though an extreme morning (lark) or evening (owl) type preference on the Morning-Eveningness questionnaire does not indicate a circadian rhythm sleep-wake disorder, those with delayed sleep-wake phase disorder tend to be extreme evening types ^{60, 61} and those with advanced sleep-wake phase disorder tend to be extreme morning types⁶².

One of the most prominent features of the difference between morning/evening types is that they exhibit different timing to patterns of sleep-wake^{50, 63, 64} Morning types have also been reported to be less flexible particularly in their rising times, than evening types⁶⁵. There is also evidence that there are alterations in the dissipation and buildup of homeostatic sleep drive in those with extreme circadian preference but a normal circadian phase (i.e. intermediate time of DLMO), that is not present in those with both an extreme circadian preference and extreme circadian phase (i.e. early or late DLMO) ^{66, 67}.

The MCTQ is considerably newer than the Morningness-Eveningness questionnaire and uses a different approach to determine circadian chronotype. While the morningness-eveningness questionnaire includes questions related to an individual’s preference of the timing of various daily activities including sleep ⁵⁰, the Munich Chronotype questionnaire estimates chronotype based on the midpoint between sleep onset and offset on work-free days, corrected for “oversleep” due to the sleep debt that individuals accumulate over the workweek⁵¹. Chronotype as measured by the MCTQ is based on the assumption that sleep timing on work-free days is highly influenced by the circadian clock. The discrepancy between work and free days, between social and biological time, can be described as ‘social jetlag’⁵⁹. The Munich Chronotype questionnaire has been taken by hundreds of thousands of people online and ‘social jetlag’ has been shown to be associated with age, body mass index and other poor health behaviors and outcomes^{51, 59, 68, 69}.

Rest-Activity Rhythms

Rest-activity rhythms are one of the most prominent outputs of the circadian system. Although the control of sleep timing and quality is a combination of the interaction between both homeostatic and circadian processes, it can be significantly influenced by behavior. So while rest-activity rhythms are commonly used to assess circadian function in both animals and humans, they are not purely a function of the circadian clock. Wrist activity monitoring in humans is a commonly used to assess the timing of rest-activity cycles as a proxy of circadian timing. Detailed recommendations for how and when to use actigraphy in the clinical setting are available from various sources ^70-73. In fact, assessing the pattern of sleep-wake over 14 days (minimum 7 days) using sleep diary or wrist actigraphy with a sleep diary is part of the diagnostic criteria for the following CRSWD’s: delayed sleep-wake phase disorder, advance sleep-wake phase disorder, non-24 sleep-wake rhythm disorder (minimum 14 days), irregular sleep-wake rhythm disorder, and shift work disorder¹. An example of why this can be helpful is depicted in the actogram of wrist activity monitoring from a patient with non-24-hour sleep-wake rhythm disorder provided in Figure 2. Of note in this example is that the rest period moves about an hour later each day, a key feature of non-24 hour sleep wake disorder. It has also been shown that the circadian rhythm of melatonin and core body temperature will also progressively delay along with the rest-activity cycles in these patients ⁷⁴.

Figure 2. An example of the rest-activity rhythm in a patient with non-24 hour sleep-wake phase disorder recorded with wrist activity monitoring.

In this figure activity level (dark tick marks) is double plotted which means that each line includes two days between mid-night to mid-night and that part of the second day is plotted on the line below. The rest-activity rhythm in those with non-24-hour sleep-wake phase disorder typically progressively delays each days so that at certain times the rest period occurs during the daytime. (From Reid KJ, Zee PC. Circadian Rhythm Sleep Disorders. In: Kryger M, ed. Atlas of Clinical Sleep Medicine. 2nd ed: Saunder, Elsevier, 2013; with permission.)

Several measures can be determined using information from rest-activity cycles that are thought to reflect the strength and timing of the circadian system. The measures derived from the rest-activity cycles have been shown to be associated with morbidity and mortality, changes in response to an intervention, and even neurodegenerative disorders ^75-78. Estimates of sleep-wake timing are just one of the “circadian rhythm” variables that can be extracted from activity records. Most commercially available wrist worn activity monitors use propriety algorithms to estimate sleep-wake, and from these you are able to ascertain sleep onset, offset and seep midpoint (clock time that represents the midpoint between sleep onset and end of the sleep offset^{39, 79}). Later sleep midpoint (typically after 5am) in particular has been shown to be associated with numerous poor cardio-metabolic health outcomes^{39, 79}. An example of wrist actigraphy from a healthy older individual is provided in Figure 3, the average bedtime and wake time for this individual are 11:40pm and 08:00 am respectively, total sleep time is about 7.5 hours.

Figure 3. An example of the rest-activity rhythm from a healthy older adult recorded with wrist activity monitoring.

Each row is a 24-hour day from 12pm to 12pm the next day. Activity levels are indicated in black and light levels in yellow. The light blue shaded area indicates the sleep period. The small blue triangles at the beginning and end of the light blue shaded sleep period are markers selected by the individual to indicate when they went to bed and when they woke up. The individual has a stable nocturnal sleep-wake period and woke up several times during the night (an example of this can be seen on day 3 as the increase in activity and light within the sleep period).

Another feature that is available in some actigraphy devices (see Figures 3 and 8), is the ability to measure light levels. While measuring light levels at the wrist has it’s limitations, and may not accurately reflect light exposure at the eye, there is evidence to suggest that light levels determined with actigraphy are associated with sleep quality, mood and health related outcomes ^80-82. By monitoring when patients are exposure to relatively bright light, clinicians are able to guide patients on how to modify behaviors to reduce light exposure at the wrong circadian time as this may be perpetuating or exacerbating their circadian disturbance. Alternatively, it can also provide information about when to increase light levels to enhance phase shifting or circadian entrainment^{61, 83, 84}

Figure 8. An example of activity, light and skin temperature levels with a wrist worn device between 3pm-10am.

Activity levels are indicated in blue, light levels in yellow and skin temperature in red. The light blue shaded area indicates the sleep period. In contrast to core body temperature, you will note that skin temperature is higher during the sleep period when core temperature is normally lower.

Other measures of circadian activity rhythms (CAR) have been developed and applied to various populations. CAR methods have used traditional cosine models ⁸⁵ and more frequently now newer expanded versions of these traditional cosine models are used to map the circadian activity rhythm to activity data^{75-77, 86-88}. CAR measures include amplitude, a measure of the strength of the activity rhythm; mesor, the mean level of activity; and the acrophase or time of day of peak activity. Another of these measures is the pseudo F-statistic which represents a measure of the robustness of circadian activity rhythm, with higher pseudo-F values indicating stronger rhythms⁸⁹.

There are also other non-parametric variables that can be determined from wrist activity monitoring that are believed to reflect rhythmic behavior ^{90, 91}. These variables include, interdaily stability (IS), intradaily variability (IV) and relative amplitude (RA) of activity values, the start times and average activity of M10 (i.e. 10 h with maximal activity) and L5 (i.e. five hours with least activity). Code to calculate these measures has been published by Blume and colleagues⁹². These measures can be improved with circadian based interventions and have been associated with dementia, diabetes, blood pressure, mood and other health related measures^{11, 90, 93}

Melatonin

Melatonin is an endogenous hormone synthesized and released by the pineal gland and the onset of melatonin production is commonly used as a marker of the timing of the endogenous circadian system ^94-96. Melatonin secretion is controlled by way of a multi-synaptic neural pathway consisting of sympathetic innervations, through the SCN, pre-ganglionic neurons and post-ganglionic fibers from the superior cervical ganglion⁹⁵. In addition to being used as a marker of the circadian system, melatonin has the ability to 1) alter the timing of circadian rhythms (phase shifting^{3, 97, 98}), and 2) help to promote sleep via its influence on the SCN, as melatonin inhibits the firing rate of SCN neurons creating a sleep permissive state ^{99, 100}.

Melatonin can be detected in plasma and saliva and its major metabolite 6-sulphatoxymelatonin in urine. Melatonin levels are most commonly measured in plasma or saliva as this allows for frequent sampling (~every 30 minutes) for more precise determination of phase. Each of these approaches comes with its own specific sampling methods and considerations, including the environment in which the sample will be collected and the presence of certain medications ^101-103 and health conditions ^{104, 105} that can impact the accuracy of the melatonin levels detected.

Melatonin production can be suppressed by light, beta- blockers and non-steroidal anti-inflammatory medications^{102, 103, 106}. In addition to impacting the levels of melatonin excretion¹⁰⁷, light exposure at certain times can also shift (phase advance or delay ^{4, 5, 107}) the timing of the melatonin rhythm and as such it is useful, when possible, to also assess the habitual light exposure patterns (see rest-activity rhythms section above) of patients before and during the sampling period. The benefits of monitoring light levels during this time are two-fold, one is that you can determine whether the measure is compromised and the other is that it may provide insight on how to target instructions for adjusting habitual light exposure, which is part of the clinical management of CRSWD ¹⁰. For example, avoiding bright light at key times that may adversely impact circadian timing and seeking out light at times that may strengthen or positively impact circadian timing are a first line of clinical management for CRSWD’s ¹⁰.

While there are several commercially available melatonin assays, the determination of melatonin levels is not a standard test performed by most clinical laboratories. Even with access to a qualified laboratory, the number of samples typically collected each day make it cost prohibitive to assay samples from a single person at a time. Currently in most clinical scenarios, there would be no more than 25-30 samples collected per person/day, even when 24-hour sampling is undertaken. There are companies that can provide relatively speedy turnaround (7-10 days) to assay samples for melatonin (the speed is a function of the volume of samples that they routinely process) ¹⁰⁸.

The most commonly used marker of circadian timing taken from serial melatonin sampling is the dim light melatonin onset (DLMO). There are several agreed upon methods for determining DLMO using plasma and salivary melatonin levels, examples of these for plasma and salivary melatonin are provided in Figure 4. For the most part the main difference for the salivary melatonin level calculation of DLMO is the absolute threshold used, while the cut off for plasma is 10pg/mL the levels of melatonin detected in saliva are generally less than in plasma¹⁰⁹ and as such the saliva absolute threshold is typically 3 pg/mL ¹¹⁰. In addition, to the dim light melatonin onset which only requires sampling for a 5-6 hours, there are other markers of the melatonin rhythm that are used when sampling covers the entire melatonin profile (typically requiring sampling for 18-24 hours). The additional measures (Figure 4) include the melatonin offset (decline in circulating melatonin levels following the end of pineal production of melatonin) and midpoint of melatonin production (midpoint between the onset and offset).

Figure 4. Illustrative examples of melatonin levels measured using 3 different methods and their associated phase estimates.

(A) 24-hour rhythm of primary melatonin metabolite 6-sulphatoxymeltonin (aMT6s) derived from urine samples collected in 2-hour bins under dim light. The fitted curve reveals a significant 24-hour rhythm with maximum levels observed between 04:00am and 08:00am (**p<0.01). (B) Salivary melatonin profile collected under dim-light conditions. The low threshold dim light melatonin onset (DLMO) was defined as either the first sample to exceed and remain above a threshold of 3 pg/mL or that was 2 SD above the mean of the first 3 baseline samples (2 SD). (C) Overnight plasma melatonin profile, plotted as a percentage of maximum (dashed line) and smoothed with a Lowess curve fit to the raw data (solid line). Some frequently used phase markers are shown: DLMO at 10 pg/mL, DLMO or dim-light melatonin offset (DLMOff) at 25% or 50% of maximum levels, the midpoint, and the termination of melatonin synthesis (Synoff). (From Benloucif S, Burgess HJ, Klerman EB, et al. Measuring melatonin in humans. J Clin Sleep Med 2008;4:66-9; with permission.)

Determination of the dim light melatonin onset or the 24-hour rhythm of 6-sulphatoxymeltonin are recommended to confirm diagnosis of various circadian rhythm sleep-wake disorders including delayed sleep-wake phase disorder, advanced sleep-wake phase disorder and non-24 hour sleep-wake rhythm disorder¹. For both advanced and delayed sleep-wake phase disorders a single measurement of circadian phase is sufficient, but for non-24 hour sleep-wake rhythm disorder measurement at 2 time points 2-4 weeks apart is recommended to observe the drift in timing of the circadian clock, a key feature of this disorder ¹.

Plasma melatonin levels are typically the gold standard for assessment of circadian rhythms for research purposes but in the clinical setting it is less practical due to the need to draw serial blood samples. Usually, serial sampling requires insertion of an indwelling catheter for the duration of the sampling period. There also needs to be access to a centrifuge and freezer to be able to process the samples quickly, to maintain sample integrity. All of these factors mean that it is not practical to measure melatonin from plasma in the home. Although a benefit of collecting plasma, saliva or urine samples in the laboratory is that the environment can be controlled to minimize the impact of light on melatonin production.

Validation of at home salivary melatonin collection for assessment of DLMO has been conducted ^{108, 111, 112}. These studies compared at home salivary melatonin collection for calculating DLMO to DLMO determined from saliva samples collected in the laboratory ^{108, 111}. Measuring melatonin levels at home from saliva requires the patient to carefully follow the instructions for sample collection. Instructions typically include remaining in dim light (preferably less than 20 lux) throughout the sampling period, not eating or drinking 10-15 minutes prior to each sample, and collecting an adequate volume of saliva. A benefit to saliva sampling is that melatonin levels in saliva are quite stable and can be stored at home until they can be returned to the laboratory for assay. Selecting the appropriate sampling duration and frequency for the patient population you wish to assess is essential. For most people with conventional sleep-wake timing, melatonin production occurs at night and on average, the DLMO occurs 2-3 hours prior to habitual bedtime ¹¹³. By knowing when sleep-wake occurs, you can predict the best time window for saliva sampling in most people. But in patients with circadian sleep-wake phase disorders who may not have regular sleep-wake schedules (i.e. non-24 hour sleep-wake phase disorder, irregular sleep-wake phase disorder, or shift work disorder) the timing of sleep-wake and therefore melatonin onset may not be as predictable and because of this sampling needs to be tailored¹¹⁴. Examples of saliva melatonin sampling from a clinical setting are provided here to illustrate a few key concepts. In Figure 5, the top panel is an example of melatonin levels from a traditional saliva sampling protocol used for the determination of DLMO. In this example saliva samples were taken at 30-minute intervals for the 5 hours prior to habitual bedtime under dim light conditions and there is a clear and consistent rise in melatonin levels, such that, the DLMO onset was determined to be at 9:00pm. The lower panel of Figure 5, depicts a longer sampling duration (22 hours) and reduced sampling frequency (hourly) that might be used if you were not sure when the DLMO might be within the 24-hour day, such as in a patent with non-24hour sleep-wake phase disorder. In this example, the DLMO is at around 4:00am.

Figure 5. Salivary melatonin levels for (A) 5-hours prior to bedtime and (B) 22-hours of sampling.

The top panel (A) represents an example from a single individual using a common method for sampling salivary melatonin levels at home. This includes 11 total samples each taken at 30-minute intervals for the 5 hours prior to habitual bedtime. DLMO was calculated using the 2 SD threshold and occurred at 8:30pm, if the 3 pg/ml threshold method is used the DLMO would be reported as 9:00pm (indicated by the arrow). The bottom panel (B) represents an example for an individual that collected samples hourly for 22 hours. DLMO for this individual occurred at 3:00am using the 2 SD criteria or at 4:00am if using the 3pg/ml threshold criteria (indicated by the arrow).

Even when you have considered all of the right factors and you have provided clear instructions to the patient you may not be able to determine the DLMO at all, an example of just such a case is provided in Figure 6. In this case, it was not possible to determine whether the timing of the sample collection resulted in simply missing DLMO altogether (DLMO occurred earlier or later) or whether the individual did not follow the instructions for sample collection (i.e. they did not dim lights during the collection period, which would suppress melatonin production^{108, 115}).

Figure 6. Salivary melatonin levels for (A) 6.5 hours prior to bedtime.

In this example it was not possible to calculate the DLMO as there was no clearly discernable melatonin levels measured. It is unclear in this case whether the DLMO was simply missed altogether or whether the individual did not follow the instructions for sample collection (i.e. they did not dim lights, which would suppress melatonin production).

Measurement of urinary 6-sulphatoxymelatonin (aMT6s) can be useful in some instances for assessment of circadian rhythm function. It is particularly useful if you want to monitor aMT6s over many days, or if you want to assess melatonin levels in infants and small children or adults with dementia. From a methodological standpoint, while urine sampling is less invasive than blood sampling, the frequency of urine sampling is less controlled, as most people cannot urinate on command. However, samples from key time periods across the day can be used to estimate a 24-hour profile of 6-sulphatoxymelatonin excretion.

For 24-hour urine collection, the subject is typically instructed to void their bladder on awakening on the day of collection, and discard the first void. For the next 24-hours, each void is collected in a separate container and marked with the time and date, with the final sample consisting of the first void the following morning. All samples are kept on ice after collection, and brought to the laboratory the following day. In the laboratory the total volume of each void is recorded, and an aliquot of each sample is frozen for further analysis. At least 4 samples are needed for analysis of circadian parameters, these include samples from the morning, afternoon, evening, and 1^st morning void the following day. 6-sulphatoxymelatonin levels need to be adjusted with total urine volume or creatinine levels^{110, 116, 117}. Twenty-four hour profiles of 6-sulphatoxymelatonin (Figure 4) can be analyzed using a cosinor analysis ¹¹⁰ to determine variables such as mesor (average of samples in the analysis), acrophase (time of peak concentration) and amplitude (difference between the acrophase and mesor).

Overnight aMT6s excretion rates or levels can also be calculated from a single morning urine sample upon waking^{110, 116, 117}. In shift workers aMT6s levels in urine from pooled samples collected during sleep and upon waking, from each sleep period over many days have been shown to be associated with sleep quality and sleep duration, such that when 6-sulphatoxymelatonin levels are high, sleep quality and duration better ¹¹⁸.

Body Temperature

In humans the circadian rhythm of core body temperature (CBT) has been used to assess the timing of the circadian system, and for many years was the most commonly used method. Typically the nadir of the core body temperature rhythm is used as a marker of circadian phase. With the discovery of melatonin and development of melatonin collection and assay methods, a shift has occurred towards melatonin as the circadian marker of choice. The reason for this shift, at least in part, stems from the masking effects of activity and meals etc. on core body temperate and that measurement of CBT was usually collected via a probe inserted into the rectum ^{119, 120}. Even with these caveats, the circadian rhythm of CBT is a useful measure of circadian rhythms, and is particularly useful if you want to determine circadian timing quickly and not have to wait for the return of assay results, which typically require a minimum 7-10 day turnaround. Furthermore, there have been technological advances that provide options for measuring core body temperature such as devices that can be swallowed and transmit real time temperature data as they pass through the digestive tract¹²¹. CBT is closely linked with sleep-wake³¹ and CBT nadir typically occurs during the latter part of the habitual sleep period in those with conventional sleep times (i.e. around 3-4 am). However, in those with a circadian rhythm sleep-wake phase disorders the CBT can be at an earlier or later clock time ⁶⁰ or even outside of the normal sleep-wake period as would be commonly seen in shift workers¹²². The profile of the CBT rhythm is typically the opposite of the melatonin rhythm. While melatonin levels are high during the biological night, CBT is low (ref). The ability to fall asleep is typically greatest when core temperature levels are declining (ref). In addition, administration of exogenous melatonin during the day when melatonin level are typically low and CBT is high can also acutely reduce CBT levels and increase sleepiness¹²⁰.

An example of a 24-hour profile of CBT from a patient with delayed sleep-wake phase disorder is shown in Figure 7; in this case the CBT nadir was at 9am. The 24-h CBT profile shown here was quantitatively described using the Cleveland regression procedure to assess the phase of the rhythm¹²³.

Figure 7. 24-hour recording of core body temperature.

This is an example of a 24-hour recording of rectal temperature from a patient with delayed sleep-wake phase disorder. Clock time (hours:minutes) is indicated on the x-axis and temperature (degrees Celsius) on the y-axis. The raw temperature data is indicated by the blue line while the pink line indicates the fitted curve for the calculation of core body temperature minimum. The calculated core body temperature is indicated by the black arrow and was at 9am. The declining portion of the CBT rhythm begins at about 3am and the rising potion of the rhythm occurs at 11am. This would make it difficult for this person to fall asleep earlier than 3am and wake before 11am and if they did wake at 11am they would likely report feeling very sleepy as the CBT is still low. (From Reid KJ, Zee PC. Circadian Rhythm Sleep Disorders. In: Kryger M, ed. Atlas of Clinical Sleep Medicine. 2nd ed: Saunder, Elsevier, 2013; with permission.)

Skin temperature can also be used as a marker of circadian timing^{124, 125}. While this method is much less invasive than measures of CBT, interpretation of temperature profiles using this method should be used with caution as ambient temperature, location of the sensors and activity level can substantially influence the measurement. Something to note about skin temperature is that it is typically higher during the sleep period (Figure 8), compared to CBT which is lower during sleep (Figure 7). There are several commercially available devices that measure skin temperature including small battery sized devices ¹²¹ that can be taped to the skin or there is a wrist worn device available that integrates skin temperature, activity and light level measurement ¹²⁶. An example of data from such an integrated device is provided in Figure 8.

The Future

Assessment of circadian rhythms in humans has been and continues to be challenging. As our knowledge about the complexities of the circadian system and the technologies to measure these systems advance, there have been exciting developments in potential new markers of the circadian system that may soon be transferable to the clinical environment.

With the discovery that almost all cells of the body contain the genetic components of the clock, and development of new technologies the methods to assess circadian rhythms have expanded. For example, we are no longer limited to measuring circadian rhythm outputs in humans but we can also measure the rhythms of the bacteria that live within them. It is thought that these bacteria may influence the function of the brain, gut and other systems within the body¹²⁷. We can also measure circadian expression of core clock and other genes in peripheral blood mononuclear cells (PBMC), a limitation of this technology has been that it still requires serial sampling for 24-hours or more. However, there are several groups working on techniques and analytical methods to limit the number of blood samples that are required to accurately determine circadian rhythms^128-133.

We also have a better understanding of how external stimuli (zeitgebers) influence the circadian system, and how the systems that relay this information to the suprachiasmatic nucleus can be altered in various patient populations. As an example pupilometry is being used to determine the responsiveness of the retina to light stimuli, while this is not a direct measure of circadian rhythms, it has implications for how the external environment interfaces with the circadian clock to impact circadian rhythms. A recent publication in those with DSWPD suggests that there is potential for the pupillary light reflex to clinically differentiate between DSWPD patients with normal vs. delayed circadian timing relative to desired bedtime, without the need for costly and time-consuming circadian assessments⁸.

Finally, it is also becoming clearer that circadian disturbance is not limited to those who meet the criteria for circadian rhythm sleep-wake disorders⁵⁹. Due to our 24/7 lifestyle and the shifting of the sleep-wake period from day-to-day and from weekday-weekends even this seemly benign level of circadian disruption if extreme enough can result in poor mental and physical health outcomes^{11, 39, 41-45, 59, 68, 134, 135}. In fact, circadian disturbance itself may be a risk factor or a marker for disease severity, or even a marker for eventual progression to disease.

Conclusion

As the role the circadian system plays in health and safety is better understood, the need to assess circadian rhythms has expanded. Even current methods of assessing circadian rhythms are not readily available to the average physician for use in the management of patients with circadian rhythm disturbance, rather they are relegated to the research environment. Translation of currently available methods of assessing circadian rhythms to the clinical environment, and development of new simple and cost effective methods for the assessment of circadian rhythms will be essential for the future of circadian medicine.

KEY POINTS.

• Assessment of circadian rhythms is likely to benefit the diagnosis, treatment and long-term management of those with circadian rhythm sleep-wake disorders.

• The dim-light-melatonin-onset is the current gold standard marker of circadian phase.

• The degree of circadian disruption maybe be a marker of severity of the disorder and direct the selection of treatment strategies.

• Development of new comprehensive tools to assess circadian rhythms across a variety of domains such as expression of molecular clocks may lead to better management and targeted treatment for circadian rhythm sleep-wake disorders.

SYNOPSIS.

Circadian rhythms are observed in most physiological functions across a variety of species and are controlled by a master pacemaker in the brain called the suprachiasmatic nucleus. In addition, the molecular machinery to generate circadian rhythms is present in most cells of the body. The complex nature of the circadian system and the impact of circadian disruption on sleep, health and wellbeing support the need to assess internal circadian timing in the clinical setting. The ability to assess circadian rhythms and the degree of circadian disruption, can help in categorizing subtypes or even new circadian rhythm disorders, and aid in the clinical management of the these disorders.