Circadian Rhythm Phase Shifts Caused by Timed Exercise Vary with Chronotype in Young Adults

University of Kentucky UKnowledge

Theses and Dissertations--Kinesiology and Health Promotion Kinesiology and Health Promotion

2019

CIRCADIAN RHYTHM PHASE SHIFTS CAUSED BY TIMED EXERCISE VARY WITH CHRONOTYPE IN YOUNG ADULTS

J. Matthew Thomas University of Kentucky, [email protected] Digital Object Identifier: https://doi.org/10.13023/etd.2019.406

Right click to open a feedback form in a new tab to let us know how this document benefits ou.y

Recommended Citation Thomas, J. Matthew, "CIRCADIAN RHYTHM PHASE SHIFTS CAUSED BY TIMED EXERCISE VARY WITH CHRONOTYPE IN YOUNG ADULTS" (2019). Theses and Dissertations--Kinesiology and Health Promotion. 64. https://uknowledge.uky.edu/khp_etds/64

This Doctoral Dissertation is brought to you for free and open access by the Kinesiology and Health Promotion at UKnowledge. It has been accepted for inclusion in Theses and Dissertations--Kinesiology and Health Promotion by an authorized administrator of UKnowledge. For more information, please contact [email protected]. STUDENT AGREEMENT:

I represent that my thesis or dissertation and abstract are my original work. Proper attribution has been given to all outside sources. I understand that I am solely responsible for obtaining any needed copyright permissions. I have obtained needed written permission statement(s) from the owner(s) of each third-party copyrighted matter to be included in my work, allowing electronic distribution (if such use is not permitted by the fair use doctrine) which will be submitted to UKnowledge as Additional File.

I hereby grant to The University of Kentucky and its agents the irrevocable, non-exclusive, and royalty-free license to archive and make accessible my work in whole or in part in all forms of media, now or hereafter known. I agree that the document mentioned above may be made available immediately for worldwide access unless an embargo applies.

I retain all other ownership rights to the copyright of my work. I also retain the right to use in future works (such as articles or books) all or part of my work. I understand that I am free to register the copyright to my work.

REVIEW, APPROVAL AND ACCEPTANCE

The document mentioned above has been reviewed and accepted by the student’s advisor, on behalf of the advisory committee, and by the Director of Graduate Studies (DGS), on behalf of the program; we verify that this is the final, approved version of the student’s thesis including all changes required by the advisory committee. The undersigned agree to abide by the statements above.

J. Matthew Thomas, Student

Dr. Jody L. Clasey, Major Professor

Dr. Melinda Ickes, Director of Graduate Studies

CIRCADIAN RHYTHM PHASE SHIFTS CAUSED BY TIMED EXERCISE VARY WITH CHRONOTYPE IN YOUNG ADULTS

______

DISSERTATION ______A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the College of Education at the University of Kentucky

By J. Matthew Thomas Lexington, Kentucky Director: Dr. Jody L. Clasey, Professor of Kinesiology and Health Promotion Lexington, Kentucky 2019

ABSTRACT OF DISSERTATION

CIRCADIAN RHYTHM PHASE SHIFTS CAUSED BY TIMED EXERCISE VARY WITH CHRONOTYPE IN YOUNG ADULTS The circadian system controls 24-hour cycles of behavior and physiology, such as rest- activity and feeding rhythms. The human circadian system synchronizes with, or entrains to, the light/dark cycle (sunrise/sunset) to promote activity and food consumption during the day and rest at night. However, strict work schedules and nighttime light exposure impair proper entrainment of the circadian system, resulting in chronic circadian misalignment. Numerous studies have shown that chronic circadian misalignment results in poor health. Therefore, therapeutic interventions that could shift circadian rhythms and alleviate circadian misalignment could broadly impact public health. Although light is the most salient time cue for the circadian system, several laboratory studies have shown that exercise can also entrain the internal circadian rhythm. However, these studies were performed in controlled laboratory conditions with physically-active participants. The purpose of this study was to determine whether timed exercise can phase advance (shift earlier) the internal circadian rhythm in sedentary subjects in free-living conditions. Fifty- two young, sedentary adults (16 male, 24.3±0.76 yrs) participated in the study. As a marker of the phase of the internal circadian rhythm, we measured salivary melatonin levels (dim light melatonin onset: DLMO) before and after 5 days of timed exercise. Participants were randomized to perform either morning (10h after DLMO) or evening (20h after DLMO) supervised exercise training for 5 consecutive days. We found that morning exercisers had a significantly greater phase advance than evening exercisers. Importantly, the morning exercisers had a 0.6h phase advance, which could theoretically better align their internal circadian rhythms with the light-dark cycle and with early- morning social obligations. In addition, we also found that baseline DLMO, a proxy for chronotype, influenced the effect of timed exercise. We found that for later chronotypes, both morning and evening exercise advanced the internal circadian rhythm. In contrast, earlier chronotypes had phase advances when they exercised in the morning, but phase delays when they exercised in the evening. Thus, late chronotypes, who experience the most severe circadian misalignment, may benefit from exercise in the morning or evening, but evening exercise may exacerbate circadian misalignment in early chronotypes. Together these results suggest that personalized exercise timing prescriptions based on chronotype could alleviate circadian misalignment in young adults. KEYWORDS: Circadian Rhythms, Chronotype, Circadian Misalignment Exercise, Phase Shift

J. Matthew Thomas

October 11, 2019 Date CIRCADIAN RHYTHM PHASE SHIFTS CAUSED BY TIMED EXERCISE VARY WITH CHRONOTYPE IN YOUNG ADULTS

By J. Matthew Thomas

Dr. Jody L. Clasey Director of Dissertation Dr. Melinda Ickes Director of Graduate Studies September 16, 2019 Date

Table of Contents

List of Tables……………….………..……………………………………..…….………vi List of Figures…………...………………………………..…………………………..….vii Chapter 1: Review of Literature……………………………….………..…………1 1.1 Overview…………………………………………………………….……1 1.2 Circadian rhythms in humans…….………………………..………...……2 1.2.1 What is a circadian rhythm?……..………………………………….2 1.2.2 Fundamental properties………..……………………………….……2 1.2.3 How are circadian rhythms generated?…………...….………….…..4 1.2.3.1 Circadian pacemaker……………...………………..………..4 1.2.3.2 Molecular mechanism…..……………….…..……...……..…5 1.2.3.3 Peripheral circadian oscillators in mammals……..…..…..….7 1.2.4 Modulation of circadian rhythms…………….…….…………….....9 1.2.5 Measuring human circadian rhythms……….…………………..…12 1.2.5.1 Melatonin………….…..…..……………………....……….12 1.2.5.2 Body temperature…………....………………………..…...14 1.2.5.3 Actigraphy…….……...…….……………………….……..15 1.2.5.4 Subjective questionnaires………...………....………….….16 1.2.6 Disruption of human circadian rhythms……………………..…….17 1.2.6.1 Shift work……………..……………………….…………..17 1.2.6.2 Circadian misalignment………....……………..………….18 1.2.6.2.1 Chronotype………………..……………….……19 1.2.6.2.2 Impact of chronotype on human performance and health…………………………………………...….……...21 1.2.6.2.3 Social jetlag…….……………..……….………..23 1.2.6.2.4 Other measures of circadian misalignment…..…24 1.2.7 Interventions targeting the circadian system……………………..26 1.2.7.1 Light therapy…………..…..………….…………….……26

iii

1.2.7.2 Time restricted feeding…………..………………….…...26 1.2.7.3 Melatonin………………………..…..………….………..28 1.2.7.4 Physical activity……………..…………..…………..…...28 1.3 Physical activity and human health……………………………………….29 1.3.1 Importance of physical activity and physical fitness…….……..…..29 1.3.2 Physical activity dose-response……...….…….…..………………..30 1.3.3 Physical activity and sleep……………...………….……………….33 1.3.4 Measurement of physical fitness ……………….…….……………34 1.3.4.1 Cardiorespiratory fitness……..…..………………………..34 1.3.4.2 Body composition……..…….…………………………….35 1.3.4.3 Muscular Fitness………....…………………………..……36 1.3.4.4 Flexibility……………..…………………………..……….36 1.3.5 Measurement of physical activity………....………………..………36 1.3.5.1 Accelerometers………..…..……...……………………….36 1.4 Physical activity and circadian rhythms…………………….……………..37 1.4.1 Phase shifting the circadian clock with exercise……..………....….37 1.4.2 Mechanisms causing exercise phase shifts…….…….…….…...…..44 1.4.3 Exercise timing and performance……………....……………….….44 1.5 Conclusion…………………………………………………….……………46 Chapter 2: Circadian rhythm phase shifts caused by timed exercise vary with chronotype in young adults…………………….…………….…………..……………...47 2.1 Abstract……………………………………………………………………..47 2.2 Introduction…………………..……………………………………………..49 2.3 Results……………………………………………………………………....50 2.3.1 Study Participants…………………………………………………….50 2.3.2 The time of exercise affects the phase of circadian rhythms in young, sedentary adults………………………………………..………….....51 2.3.3 Chronotype is associated with phase shifts in evening exercisers...... 55 2.4 Discussion…………………………………………………………………..57

2.5 Methods……………………………………………………………………..61 2.5.1 Participants……………………………………………………………61 2.5.2 Study protocol…………………………………………………………61 2.5.3 Chronotype questionnaires……...……………………………………..62 2.5.4 Anthropometric and body composition measures…….……..………...62 2.5.5 Actigraphy……………………………...…………………………...…63 2.5.6 Circadian phase measure: dim light melatonin onset (DLMO)...….….63 2.5.7 Maximal graded exercise test……………..………………………..…64 2.5.8 Exercise intervention……………..…..…………………………….....65 2.5.9 Phase shifts……………………..…………………………………..…65 2.5.10 Statistical analysis……………...………………………………….…66 2.5.11 Study Approval………………….....……………………………...…66 2.6 Supplemental information……..……………………………………………...67 Chapter 3: Expanded Methodology and Procedures……………………………………..70 3.1 Study approval………………………………………………………………..70 3.2 Subjects recruitment………..………………………………………………....70 3.3 Protocol………………………………………………………………………..71 3.3.1 Anthropometric measures and body composition...………………….….71 3.3.2 Cardiorespiratory fitness……………………………………………...….72 3.3.3 Circadian and sleep measures…..………………………………………..73 3.3.3.1 Dim light melatonin onset…..…………………………………...73 3.3.3.2 Chronotype………………………………………………………78 3.3.3.3 Actigraphy and sleep logs…………………………………….….78 3.3.4 Exercise intervention………………………….………………………….80 3.3.5 Statistical analysis……………………………………………………..…81 3.4 Excluded and withdrawn subjects………………………………………….….81 References………………………...……………………………………………………..84 Curriculum Vitae…………………………………………..…………………………...102

List of Tables Table 1.1 Summary of research investigating circadian phase shifts with exercise……..39 Table 2.1 Characteristics of study participants ……………….…………….………..….52 Table 2.2 Circadian and sleep characteristics of study participants……………….….....54 Table 2.3 Statistical model to assess covariates ………………………...………..….…..54 Table 2.4 Model of morning and evening exercise on earlier and later chronotypes ..….56 Table 2.5 Supplemental table - baseline characteristics of study participants by sex .….67

List of Figures Figure 1.1 Components of a circadian system………….…………………………………3 Figure 1.2 Hierarchical organization of the mammalian circadian system……….…...….8 Figure 1.3 Example phase response curve to light exposure…………………..……...…10 Figure 1.4 Melatonin rhythm with dim light melatonin onset (DLMO) indicated…..…..13 Figure 1.5 Circadian misalignment measured as the phase relationship between DLMO and sleep onset ……………………….………………………………………..………...25 Figure 1.6 Phase advances of the internal circadian rhythm could improve circadian alignment in late chronotypes……………………………..…………………..…………38 Figure 2.1 Participant recruitment…………………….…………………………………51 Figure 2.2 Morning exercise advances the phase of the internal circadian rhythm…..….55 Figure 2.3 Phase shift is correlated with internal phase in evening exercisers………..…56 Figure 2.4 The effects of timed exercise on phase shift depends on chronotype………..57 Figure 2.5 Study protocol………………………………………………………..………62 Figure 2.6 Supplemental figure. Chronotype is normally distributed…………..…….....68 Figure 2.7 Supplemental figure. MEQ is associated with sleep fragmentation during exercise intervention……………………………………………………………..……....68 Figure 2.8 Supplemental Figure. Baseline DLMO is associated with mid-sleep during exercise intervention…………………………………….……………………………….69 Figure 3.1 Recruitment flyer…………………………………………………...………...71 Figure 3.2 DLMO collection instruction document………………...…………………....75 Figure 3.3 DLMO setup for circadian phase measurement………………...…………....76 Figure 3.4 Dim light melatonin onset calculation……………………...………………...77 Figure 3.5 Sleep logs………………………………………………………...…………...79 Figure 3.6 Excluded subject DLMOs…………………………………………..…..……83

vii

Chapter 1. Review of Literature

1.1 Overview

Exercise has many well-established benefits to physical and mental health (1-6). Additionally, many aspects of exercise prescription, including intensity, duration, frequency, mode, volume, and progression, have been studied to determine the optimal approaches to benefit health (2, 7, 8). However, the proper time of day of exercise to achieve maximal health benefits has not been elucidated. Determining the ideal time of day to exercise could have a broad and dramatic impact on human health.

The effects of timed exercise on health are likely mediated by the circadian system. Circadian rhythms are 24-hour cycles of physiology and behavior that are synchronized with the daily cycles caused by the rotation of the Earth. Humans have evolved with the 24-hour light/dark cycle of sunrise and sunset. The circadian system synchronizes with this 24-hour cycle and regulates rhythms of behavior and physiology so that they anticipate and are synchronized with the daily availability of resources (9). Humans are diurnal, meaning the circadian system promotes activity and arousal during the day and rest at night.

Modern society has experienced a massive departure from reliance on environmental cycles to shape daily lives and has limited exposure to natural light/dark cycles. The advent of electric lighting, as well as the use of TVs, tablets, and smartphones, has also resulted in light exposure at night. These changes in lifestyle have paralleled the increases in the prevalence of obesity and sedentary behavior.

Numerous studies have shown that circadian disruption has negative effects on health. For example, shift work, which chronically disrupts the circadian system, increases the risk of cancer, obesity and cardiovascular disease (10-15). Why is the circadian system important to human health? In this review I will first discuss the fundamental properties of circadian rhythms, the discovery of circadian clocks in tissues, methods for measuring circadian rhythms, and the consequences of disrupting the circadian system. Next, I will discuss the health benefits of regular physical activity, the proper exercise dose to improve health, and the importance of physical fitness, as well as methods of measuring

1 fitness. Finally, I will discuss the current body of evidence supporting an effect of exercise on the circadian system.

1.2 Circadian rhythms in humans

1.2.1 What is a circadian rhythm?

Circadian rhythms are approximately 24-hour cycles of behavior, physiology, and gene expression. Circadian rhythms have evolved to entrain to environmental cycles but can continue to oscillate even without environmental input. Observations of circadian rhythms go back as far as the fourth century B.C. (16). However, it wasn’t until the 18th century when the first experiment was performed, demonstrating that Mimosa plants have circadian rhythms of leaf movement (17). Two hundred years later, in the middle of the 20th century, there was a surge of research in the field of circadian rhythms. Work by Colin Pittendrigh, Jorgen Aschoff, and Franz Halberg (who coined the term circadian rhythms in 1959; cira-diem, latin for “about a day”) was pivotal in developing the field of circadian biology. In one of Pittendrighs’ early works (18) he proposed that a circadian system must have 3 essential components: 1. rhythmi c environmental input, or zeitgeber (German for time giver); 2. an endogenous self-sustained oscillator; and 3. output rhythms of physiology and behavior (Figure 1.1). This is now the dogmatic model of the circadian system.

1.2.2 Fundamental Properties

There are 3 fundamental properties of a circadian rhythm. These properties distinguish circadian rhythms from simple responses to external stimuli and demonstrate that it oscillates autonomously.

First, a circadian rhythm must be endogenous, such that the rhythms continue to cycle with a ~24-hour period when placed in constant conditions without external time cues. This is called the free-running rhythm. The free-running periods of most organisms are near, but not exactly 24 hours. The fact that circadian rhythms are not exactly 24 hours is evidence that circadian rhythms are generated endogenously and are not driven by an environmental stimulus (since environmental stimuli are often exactly 24 hours). An internal rhythm, which continues to oscillate in constant conditions, has been observed in

2 many organisms spanning many phyla. In the 1950s researchers showed that several different organisms presented free-running rhythms, even after being in constant conditions since birth or for multiple generations (18). In the past several decades, research on human circadian rhythms has involved controlled study designs aimed at observing human circadian rhythms with no time cues or environmental stimuli. These studies have taken place in caves, bunkers, and controlled lab conditions with the goal of limiting external perturbations that can effect circadian rhythms (19, 20).

Figure 1.1 Components of a circadian system. A circadian system has 3 essential components: A) cyclic input (zeitgeber), which is most often the light-dark cycle in humans; B) a self-sustained oscillator that can keep time in the absence of input, but is also capable of entraining to the cyclic input; and C) Output rhythms that are driven by the oscillator and entrained to the cyclic input.

The second fundamental property of a circadian rhythm is that it must entrain to environmental stimuli that have a 24-hour rhythm. The ability of internal rhythms to entrain to environmental inputs is critical for the adaptation to the 24-hour cycle. Entrainment is the process in which a circadian rhythm (which is near, but not exactly 24 hours) adopts the period of an input stimuli (usually the 24-hour light/dark cycle) and develops a stable phase relationship. Two distinct models of entrainment, parametric and non-parametric, have been proposed by the founders of the circadian biology field, Aschoff and Pittendrigh, respectively (21, 22). The non-parametric model, involving

3 instantaneous shifts to the internal rhythm, is the most well-established model and will be discussed in a later section.

The third fundamental property is that a circadian rhythm must be temperature compensated. This means that the period of the rhythm remains constant over a physiological range of temperatures. Biological reactions are sensitive to temperature. If a circadian rhythm was not temperature compensated, the period of the rhythm would vary depending on the external temperature. Early evidence of temperature compensation investigated ectothermic organisms, including bees, crabs, and Drosophila (23, 24). More recently, temperature compensation has been observed in neural and peripheral tissues cultured from endotherms (within normal physiological temperature ranges), including the suprachiasmatic nuclei (SCN) and fibroblasts (25). Very little evidence is available to demonstrate temperature compensation in humans. However, one study did show temperature compensation in fibroblast samples from one subject (26).

1.2.3 How are circadian rhythms generated?

In the 1950s, the 3-component model of the circadian system proposed by Colin Pittendrigh (Figure 1.1) was quite controversial (18). Specifically, the existence of a self- sustained endogenous oscillator was a point of contention. A popular hypothesis at the time explained daily rhythms with an “hourglass model” involving a substance that accumulated during the night and was depleted during the day (27). Another obstacle at that time was that the anatomical location of the hourglass or endogenous oscillator was unknown. In the following paragraphs, I describe the discovery of circadian oscillators in mammals, and how these discoveries supported Pittendrigh’s model, containing a self- sustained oscillator.

1.2.3.1 The Circadian Pacemaker

Prior to the 1970s, the location of the tissue responsible for generating 24-hour biological rhythms was a mystery. However, researchers knew that a rostral portion of the hypothalamus controlled the rhythmic estrous cycle (28), and it therefore was a candidate for the circadian pacemaker. Fredrick Stephan and Irving Zucker (29) lesioned an area of the hypothalamus called the suprachiasmatic nucleus (SCN) and measured drinking and

4 activity in rats. Whereas control rats consumed 95% of their water during the dark phase, SCN-lesioned rats completely lost their nocturnal drinking rhythm. Nocturnal wheel running activity was also significantly reduced, thereby disrupting the activity rhythm.

Moore and Eichler (30) then provided further information about the ability of the clock to synchronize with the light/dark cycle. They investigated the role of retinal projections in modulating the 24-hour rhythm of corticosterone in rats. They found that the corticosterone rhythm was not entrained to the light-dark cycle when the retinohypothalamic tract was severed. The corticosterone rhythm was also destroyed when the SCN was lesioned. Thus, it was concluded that the SCN was responsible for generating neuroendocrine rhythms and the retinohypothalamic tract was essential for entraining 24-hour rhythms with the light/dark cycle.

Even with this considerable evidence suggesting the SCN was the mammalian pacemaker tissue, further evidence was warranted. Therefore, Michael Menaker and colleagues (31) sought to determine if transplanting SCN tissue to an SCN-lesioned animal could restore circadian rhythms in the recipient animal. Wild-type and tau mutant hamsters were the perfect model organisms because of their distinctly different period lengths. Wild-types had a 24-hour period and homozygous tau mutants had a 20-hour period. After SCN ablation the hamsters became completely arrhythmic, as evidenced by their arrhythmic locomotor activity. Fetal donor SCN tissue was implanted in each of the hamsters. Following the implant, the locomotor activity of the hamsters returned. Incredibly, the hamsters exhibited a period that was similar to the genotype of the donor SCN, therefore providing evidence that the SCN is autonomous in the generation of mammalian circadian rhythms. Also, this experiment showed that the SCN tissue encoded the period of circadian rhythms and regulated output rhythms.

1.2.3.2 Molecular Mechanisms

Even before the discovery of the SCN, the first evidence of the molecular mechanisms of the circadian clock came from Drosophila (32-34). After hatching, Drosophila emerge as maggots, and then develop into pupae. Emergence from the pupae (termed eclosion) into an adult form occurs at dawn and is a bona-fide circadian rhythm. Konopka and Benzer (35) sought to determine if genes could encode the circadian rhythm of eclosion. To

5 determine this, they mutagenized Drosophila and identified mutants with abnormal eclosion rhythms. Following observation of eclosion rhythms from thousands of flies, 3 types of mutants were observed. One mutant was arrhythmic and emerged at any time of day or night. One had a short period of approximately 19 hours and emerged hours before dawn. The third mutant type had a long period of about 28 hours and emerged hours after dawn. To locate the position of the mutation, recombination experiments were performed. The three types of mutants were all products of alterations to a single gene that affected both eclosion and locomotor rhythms. This was the first evidence of a gene that was responsible for the period of the circadian oscillator. Thus, Konopka and Benzer named it the period (per) gene. The per gene was later cloned, which was a major discovery that earned three circadian biologists, Drs. Rosbach, Hall, and Young, the 2017 Nobel Prize in Physiology or Medicine.

Konopka and Benzer discovered a major component of the clock, but their studies did not address how the per gene regulated circadian rhythms. To investigate this, Hardin and colleagues (34) exposed wild-type flies to a cycle of 12 hours of light and 12 hours of dark prior to freezing individual flies at various circadian times. They found that per mRNA peaked at the time of lights-off and was at its lowest in the middle of the light period. They also observed rhythmic fluctuations of per mRNA in mutant flies with abnormal period lengths. However, in mutant flies they noticed that the mRNA fluctuations were either earlier or later compared to the wild-types. Even in 24-hour dark conditions per mRNA expression was rhythmic. In fact, the mRNA fluctuations were identical in period length to the activity rhythms of the Drosophila. This study concluded that rhythmic fluctuations in per mRNA underlie fluctuations in PER protein and possibly overall circadian pacemaker rhythmicity. They further postulated that the per gene product regulates its own transcription by negative feedback and drives the clock function and the output rhythms.

Although the discovery of the period gene in Drosophila was integral to understanding the circadian clock, much more information was needed, particularly in mammals. An explosion of genetic findings in the mammalian circadian system began with the discovery of the Clock gene (36). Similar to the approach used by Konopka and Benzer in

Drosophila, Takahashi et al. (36) mutagenized mice and measured activity rhythms. The Clock mutant mice had a long period and became arrhythmic in constant darkness. The position of the Clock gene was also mapped on the mammalian chromosome. Following this, the other components of the mammalian molecular clock were discovered.

The current model of the molecular timekeeping mechanism of circadian clocks in mammals is a transcriptional/translational feedback loop of gene and protein expression (37-40). Two transcription factors, CLOCK and BMAL1, heterodimerize and drive the transcription of Per and Cry mRNA. After translation of these mRNAs into protein products they dimerize and are then able to enter the nucleus. Once in the nucleus, PER and CRY proteins inhibit the transcription factor activity of CLOCK and BMAL1. It takes approximately 24-hour for one cycle of this transcription/translation feedback loop. Other mechanisms, such as post-translational modifications, refine the molecular feedback loop to modulate period length and to permit entrainment to environmental cycles.

1.2.3.3 Peripheral circadian oscillators in mammals

Research at the turn of the 21st century expanded our understanding of the components of the mammalian circadian system. By this time, thousands of circadian rhythms in behavior, physiology, biochemistry, and gene expression had been discovered. It seemed unlikely that one solitary circadian oscillator in the SCN could directly regulate all these output rhythms. Several lines of evidence in rodents, and other species, suggested that peripheral tissues may contain self-sustaining circadian oscillators (see (41) for review). If a peripheral tissue contained a self-sustained clock, then its rhythms should persist in a dish. (42, 43). Yamazaki et al. (42) to ok advantage of the recent cloning of the Per1 gene and promotor to generate a reporter animal. By linking the firefly luciferase gene to the mouse Per gene promoter, a transgenic rat model was created. This allowed the visualization of the rhythmic activity of this promoter in peripheral tissues in a dish by measuring bioluminescence. This technique allowed for the observation of circadian rhythms in any tissue ex vivo. Researchers observed that in addition to the SCN, peripheral tissues (e.g. lung, skeletal muscle, liver) exhibited robust circadian rhythms of Per1 promoter-driven bioluminescence when they were cultured in a dish. In addition,

7 the rhythm of each tissue peaked at a unique time, relative to the SCN rhythm, that was similar for all rats in the study, indicating a stable phase relationships between each tissue and the SCN. Also, large shifts (6 hours) in the light/dark cycle resulted in immediate shifts of the SCN, but slower shifts in peripheral rhythms. In a similar study (43), even SCN lesioned mice had intact rhythms in peripheral tissues. Importantly, SCN lesioning resulted in large differences in the phase of peripheral rhythms in each tissue, suggesting that the SCN coordinates the phases of peripheral tissues.

Figure 1.2 Hierarchical organization of the mammalian circadian system. A circadian pacemaker (SCN) receives light information and entrains to the light dark cycle. The SCN then coordinates the phases of peripheral oscillators. These oscillators then control local circadian output rhythms.

Many studies have contributed to the development of the hierarchical model of the mammalian circadian system (42-44). This model includes a circadian pacemaker (SCN),

8 which receives light information and entrains to the light dark cycle. The SCN then coordinates the phase in peripheral oscillators (Figure 1.2). The SCN and peripheral oscillators then drive their respective output rhythms.

1.2.4 Modulation of circadian rhythms

The circadian system evolved to entrain to natural environmental cycles. Over the course of a day an organism is exposed to many stimuli that could contribute to entrainment. The cycle of light (sunrise) and dark (sunset) has long been considered the most salient entrainment cue for the SCN in mammals (9). This so-called photic stimulus has been well-studied. Decades ago researchers began to plot the response of various physiological and behavioral rhythms to timed light stimuli (21, 45). These plots came to be known as Phase Response Curves (PRC), which are graphical representations of phase shifts as a function of time, unique to each organism and each stimulus. Currently, PRCs exist for many organisms, nocturnal and diurnal, including humans (46-48).

Constructing a PRC is done by first releasing the organism into constant conditions (i.e., there are no rhythmic stimuli in the environment), which is typically constant darkness for rodents (47). Then light pulses are applied at different times, and the resultant shift in output rhythms are observed. Several studies in humans have used the circadian rhythm of melatonin production as an output rhythm for a PRC. When PRCs are graphed, the x- axis represents the time of the stimulus, relative to a given phase marker from a physiological rhythm. Previous human studies have plotted the number of hours after melatonin onset or hours since core temperature minimum to display the time of a stimulus relative to an individual’s internal rhythm. By calculating the change in timing of a phase marker, a phase shift is measured. The y-axis represents the magnitude of the phase shift in the output rhythm (Figure 1.3). By convention, a positive value denotes an advance in the phase and a negative value denotes a delay in the phase. The shape of the PRC (i.e., the maximal phase advance or delay) shows the range of entrainment for the pacemaker to a given stimulus. There are limits to entrainment, which are determined by the intensity and duration of the stimulus, the circadian period of the organism, and the shape of the PRC to that stimulus.

Figure 1.3 Example phase response curve to light exposure. Light pulses in the early and middle of the night, near the temperature minimum, will cause phase delays while early morning light exposure will cause phase advances (49-51).

PRCs provide us with an understanding of how we and other organisms entrain to the 24- hour daily rotation of the earth. Two modes of entrainment have been proposed: parametric and non-parametric (47, 52). Parametric entrainment is based on the premise that continuous light exposure acts to compress or extend the free running period and thereby entrain the pacemaker to the environment (45). Non-parametric entrainment is based on the premise that the circadian pacemaker is entrained by exposure to discrete light pulses. When these light pulses are delivered during phases when the circadian pacemaker is capable of shifting, it will advance or delay its timing to entrain to the external cue. Although circadian entrainment likely involves contributions of both parametric and non-parametric entrainment, non-parametric entrainment has been well- studied in recent decades. In support of non-parametric entrainment, several studies have shown that humans and other organisms are able to stably entrain to brief pulses of light (skeleton photoperiods) throughout the day (21, 51, 53).

For adult humans who have a period slightly greater than 24 hours, the pacemaker must experience a daily phase advance in order to entrain to the light-dark cycle (54). In humans, a light pulse during the early part of the night will phase-delay the circadian clock, while a light pulse during the late part of the night (early morning) will phase- advance the circadian clock (Figure 1.3) (49-51). Therefore, exposure to early morning light is advantageous to humans because it helps them entrain to the environmental light- dark cycle.

The entrainment of the circadian clock to environmental stimuli regulates the magnitude of circadian misalignment. If an individual is stably entrained with a small phase angle to the light-dark cycle (i.e., the sleep-wake cycle is aligned with the dark-light cycle), then the individual has minimal circadian misalignment. This is important because circadian misalignment is associated with insulin resistance, obesity, metabolic syndrome, and systemic inflammation (11, 13, 55-57) (see further discussion below). Thus, phase response curves can be used to understand how we anticipate our circadian rhythms will respond to a given stimulus. For example a condition called delayed sleep-wake phase disorder is associated with reduced light exposure during a critical advancing window of the PRC to light (58). Since the human period is usually slightly longer than 24 hours, this lack of daily phase advance may exacerbate the delayed sleep/wake rhythm. Considering the numerous health consequences of circadian misalignment (discussed below), it is important to understand the precise effect of photic and non-photic time cues throughout the 24-hour day.

Although many studies have investigated the impact of photic stimuli on circadian rhythms, non-photic cues can also phase shift the circadian system. For example, there is mounting evidence suggesting an effect of food intake, exogenous melatonin, and physical activity on human circadian rhythms (59-63). Although it is clear that non- photic cues can alter circadian rhythms, questions still remain regarding the optimal timing of these stimuli to improve entrainment to the light/dark cycle. Regularly scheduled non-photic stimuli could be an attractive intervention for late chronotypes and others experiencing circadian misalignment.

1.2.5 Measuring Human Circadian Rhythms

Studies in rodents and non-human primates have established that the SCN is the circadian pacemaker (64, 65). The SCN is also believed to be the circadian pacemaker in humans, although it is not possible to perform lesion or other experiments to validate this hypothesis. Therefore, in humans, the internal circadian timing, or phase, is measured by assessing rhythms that are presumably under the direct control of the SCN. These include core body temperature and hormone (e.g., cortisol, melatonin) rhythms. Though a multitude of rhythms exhibit circadian oscillations, few are appropriate for controlled investigations. Many rhythms (e.g. activity, heart rate, insulin sensitivity, and cortisol) are influenced greatly by external factors. When stimuli appear to affect a rhythm without actually altering the pacemaker, it is called “masking”. Previous research has attempted to limit the masking effect by conducting human studies in controlled conditions. One experimental condition used in many human circadian studies is called constant routine (66, 67). During the constant routine, subjects remain awake in a room in constant dim light (temperature and humidity are also constant), in a supine position, with isocaloric meals provided at regular intervals (e.g. every hour) for about 48-hours. This experimental method has been used for decades in human circadian studies. However, the highly controlled laboratory conditions make it difficult to apply any findings to free- living conditions. Circadian rhythms can also be measured in free-living conditions. Two well-established physiological markers of the internal phase are melatonin secretion and core body temperature. These measures, as well as others that are ideal free-living measures of human circadian rhythms, are discussed below.

1.2.5.1 Melatonin

The human circadian pacemaker directly controls the timing of the secretion of melatonin from the pineal gland (68). The rhythm of melatonin typically peaks during the night for people with normal sleep/wake activity (Figure 1.4). It is also a bonafide circadian rhythm since it persists in the absence of light (or in constant dim light conditions). The melatonin rhythm, measured in plasma, urine, or saliva, is a well-established method for measuring human circadian phase (68-71). Light suppresses melatonin levels. For this reason, experts have agreed that measurements of melatonin measured for circadian

12 purposes should be conducted in dim light of <30lux (71). When measuring the melatonin rhythm, any time point can be used to assess the endogenous phase. However, because melatonin onset occurs prior to sleeping, this time point is commonly chosen as the phase marker. Dim light melatonin onset (DLMO) represents the synthesis and secretion of melatonin. Using linear interpolation, DLMO is defined as the time when melatonin concentration exceeds a specific threshold. Previous studies have defined DLMO using a variety of thresholds (e.g., 3pg/ml, 4pg/ml, 5pg/ml, 10pg/ml, mean + 2SD of daytime values, 20% of peak, 25% of peak, 40% of peak) (70-74) depending on the biological fluid (saliva, blood, or urine) that is being collected. Often 4pg/ml is used as the threshold for DLMO in saliva (73, 75). Previous studies have found that DLMO is strongly associated with sleep timing (73, 76). Specifically, DLMO occurs approximately 2 hours prior to bedtime for young adults (Figure 1.4) and is significantly correlated with mid-sleep and wake time on free days (days with no vocational responsibilities) (69, 76).

Figure 1.4 Melatonin rhythm with dim light melatonin onset (DLMO) indicated.

Posture is an important factor to consider when measuring melatonin. Postural changes have been shown to alter plasma and salivary melatonin levels (77). For this reason, extraneous movement must be minimized during the DLMO assay. This is particularly important just prior to collecting each sample. Subjects are typically asked to refrain from caffeine, chocolate, bananas and alcohol on the day (if not longer) of the melatonin measure (75, 76). In addition, nonsteroidal anti-inflammatory drugs (NSAIDs) have been shown to suppress melatonin and should be excluded up to 72 hours before assessment (78). While most studies have found that there was not an effect of menstrual cycle on the melatonin rhythm, one study showed a delay in melatonin onset during the luteal phase (79-82). However, the conflicting result could have been due to the large differences in light exposure between study protocols (range: 50 to >1000 lux).

1.2.5.2 Body Temperature

Oscillations in core body temperature occur every ~24 hours. In free-living humans who have normal sleep/wake activity, core body temperature peaks around 14:00-20:00 and begins to rise just before waking, thereby anticipating and preparing the individual for waking. Previous studies in rodents have demonstrated that the core body temperature rhythm occurs as a direct output of the SCN (83, 84). Many studies in humans have established that the rhythm of body temperature persists in controlled conditions, limiting external time cues. Nathaniel Kleitman conducted one of the first studies to demonstrate the human circadian rhythm of body temperature. In 1938, Kleitman and his graduate student lived in Mammoth Cave for several weeks under a modified 28-hour day (19). Each day consisted of 19 hours of wakefulness and 9 hours of darkness in bed. The researchers observed robust circadian rhythms of oral temperature.

The traditional method for measuring core body temperature in humans is by an indwelling rectal probe, which requires that the subject remain in the lab for the duration of the experiment. Laboratory measures of the core temperature rhythm often take place in a constant routine protocol since core body temperature is sensitive to masking effects. However, the restriction of sleep (which is enforced in the constant routine) can reduce the amplitude of the temperature rhythm; therefore some researchers have suggested that

14 the core temperature rhythm measured during constant routine does not truly represent the endogenous rhythm (85, 86).

Another method to measure core body temperature is by gut temperature thermistor pills. This method can be used in free-living subjects. These temperature sensors are swallowed and collect core temperature information as the sensor passes through the digestive tract. A previous study compared rectal and gut temperature measurements and showed that they had narrow limits of agreement, while axilla temperature was very different from rectal temperature and exhibited greater limits of agreement (87). Disadvantages to using thermistor pills are that they are expensive, and often multiple pill ingestions are required due to reduced transient time and evacuation from the bowels. In addition, the subjects must wear a small transmitter around the abdomen to collect data from the sensor.

The Thermochron iButton® has been used to non-invasively measure the circadian rhythm of skin temperature in free-living subjects (88, 89). The iButton is a small metal device approximately the size of a quarter which contains a temperature sensor and a battery. The iButton is usually fixed to skin (such as the inner wrist) using medical adhesive tape (90). The iButton allows ambulatory monitoring of the skin temperature in the periphery. However, it is important to note that peripheral skin temperature rhythms are out of phase with the core body temperature rhythm, because heat is lost via heat exchange from the skin due to peripheral blood flow. Thus, iButton measurements do not measure core body temperature and therefore do not report the direct output of the internal circadian clock. Methodological issues associated with the iButton include a lack of consensus on proper skin location, and that clothing and ambient temperature affect the reliability of the data (90).

1.2.5.3 Actigraphy

One of the most well-established circadian rhythms is that of locomotor activity. The behavioral rhythm of locomotor activity has been documented in nocturnal and diurnal organisms, including humans (91, 92). In individuals with a normal sleep/wake rhythm, activity is consolidated to the daytime. Activity rhythms under free-living conditions are often measured with actigraphy. Actigraphy is the measurement of activity collected via

15 accelerometers, which are devices that detect human motion in the form of accelerations, often in three planes (tri-axial). This is a non-invasive and continuous way to detect movement. Actigraphy devices can be worn on the wrist or on a waist belt above the hip. Since the 1970s developments in actigraphy devices have produced devices with long battery life and high memory storage (greater than 2 weeks). A strength of using actigraphy is that it allows long, continuous recording under free-living conditions with minimal participant burden. For example, one previous study continuously collected actigraphy data for more than 1.5 years (92).

Actigraphy can also be used to indirectly detect sleep, assess sleep quality, and determine the sleep/wake rhythm. Actigraphy is a free-living alternative to polysomnography (PSG), which is a well-established laboratory method for measuring sleep. Many commercially available actigraphy devices have been validated by PSG (93). For measures of sleep, accelerometers should be worn on the wrist where they are sensitive to movements of the extremities. It is also helpful to have participants keep recordings of sleep/wake times and removal of the device to explain missing data or aid in data processing. Also, many accelerometer devices measure light exposure which is often an important measure in circadian studies. Previous studies have shown that sleep phase measured by actigraphy (e.g. sleep onset, mid-sleep) are highly correlated with internal circadian phase (94).

1.2.5.4 Subjective Questionnaires

Although many well-established physiological and behavioral methods have been used to measure circadian rhythms in humans, questionnaires have also been developed to assess circadian rhythms. Several well-established questionnaires have been used in previous studies that query an individual’s preferred timing of sleep, food intake, and physical and mental activities, called chronotype (95-97). Chronotype is thought to be a direct output of the internal circadian rhythm.

The first questionnaire, the Morningness-Eveningness Questionaire (MEQ), was developed in the mid-1970s and was designed to assess an individuals’ ideal time of sleep and activity throughout the 24 hour day. Another questionnaire, developed more recently, is the Munich Chronotype Questionaire (MCTQ). The MCTQ queries the time of sleep

16 and wake for the previous month, assessing actual sleep behavior. There are many benefits to including chronotype questionnaires in free-living research studies, as they are free and strongly correlated with measures of circadian phase (98, 99). More details regarding measuring chronotype will be discussed below.

1.2.6 Disruption of human circadian rhythms

The circadian system presumably evolved to coordinate behavior and physiology with environmental cycles such as sunrise/sunset (9). Modern society has seen a massive departure from reliance on environmental cycles to shape daily lives. Exposure to natural light/dark cycles that synchronize the SCN clock to the environment is limited. This has resulted in a society that is living in a chronic state of circadian disruption. Below I review the evidence that circadian disruption is negatively impacting human health.

1.2.6.1 Shift Work

Shift work, or working outside of standard daytime hours, has a tremendous impact on our circadian system. Permanent night, or rotating shift work, interrupts the finely-tuned coordination of the circadian system with the environment (sunrise/sunset). Epidemiological studies have found that shift work increases the risk of obesity, cardiovascular disease, metabolic syndrome, and certain types of cancer (100-103). An elevated risk of breast cancer has been well-established in several studies, including airline cabin crews as well as other shift working populations (104, 105). Nakamura et al. (106) found that rotating shift work was associated with a higher risk of coronary heart disease, characterized by higher levels of serum cholesterol and abdominal adiposity.

Laboratory studies have also investigated the acute effect of shift work-like circadian disruption. Morris and colleagues (107) conducted a study in which subjects participated in an 8-day circadian misalignment protocol (they had a 12-hour shift of their sleep/wake schedule similar to night shift workers). This shift work-like schedule increased 24-hour systolic and diastolic blood pressure as well as inflammatory markers including CRP, TNF-α, resistin, and IL-6. Also, the magnitude of the reduction (“dipping”) in the systolic blood pressure rhythm was reduced. The loss of the “dip” in blood pressure during sleep is an independent predictor of adverse cardiovascular events and all-cause mortality

(108). Findings from the study by Morris et al. suggested that circadian misalignment is a key mechanism causing the increased risk of cardiovascular disease in shift workers.

A protocol called forced desynchrony is often used in laboratory-based circadian studies to separate the masking effects of sleep on the circadian rhythm. For example, a common protocol requires a subject to live multiple days on reoccurring 28-hour days. Since the period of the human circadian pacemaker is approximately 24 hours, this day length is outside of the range of entrainment. Thus, the endogenous rhythm will therefore desynchronize from the imposed light/dark or sleep/wake schedule and presumably oscillate at its intrinsic rhythm. A study by Scheer and colleagues (109) used the forced desynchrony protocol to investigate the metabolic and cardiovascular consequences of circadian misalignment in 10 subjects. Subjects in this protocol were exposed to recurring 28-hour days and ate 4 isocaloric meals each day. Hourly, plasma leptin, insulin, glucose, and cortisol were measured. On the fourth day of the protocol subjects were misaligned by 12 hours with their baseline. Compared to when they were aligned, misalignment resulted in a 17% decrease in leptin, a 6% increase in glucose, and a 22% increase in insulin. Nearly 40% of subjects exhibited a prediabetic or diabetic 2-hour postprandial glucose response.

Because sleep restriction often accompanies circadian misalignment in free-living conditions, one study investigated the combination of forced desynchrony with sleep restriction on glucose metabolism (110). They found that circadian misalignment combined with sleep restriction induced a 15% increase in postprandial glucose and 27% decrease in postprandial insulin (110).

1.2.6.2 Circadian misalignment

Circadian disruption has been defined in many ways, but in general it refers to living against your internal rhythm. One term, circadian misalignment, has been defined as two or more rhythms with an abnormal phase angle relative to each other (111). Circadian misalignment can occur at many different levels, including misalignment between the circadian pacemaker and the light/dark cycle, or between the circadian pacemaker and sleep timing, or between rhythms in peripheral tissues and the circadian pacemaker. Circadian misalignment is a chronic issue that occurs in everyday life. The next sections

18 discuss methods of assessing circadian misalignment in free-living conditions and how circadian misalignment affects human health.

1.2.6.2.1 Chronotype

Years ago, in the absence of artificial light, the circadian pacemaker and the sleep/wake cycles were tightly regulated by the environmental light/dark cycle. Modern society and the advent of electric lighting have resulted in an extended period of illumination into the night, and large variability in sleep/wake schedules among individuals (112). For example, a preferred sleep onset of 3:00 AM or later has been reported in approximately 8% of the population (113).

Subjective questionnaires that query preferred timing of daily activities (e.g. sleep, wake, physical activity, food intake) can be used to assess an individual’s entrainment to the 24- hour day, called chronotype. The Horne and Osterberg questionnaire, also referred to as the Morningness Eveningness Questionnaire (MEQ), assigns chronotype based on a subjective rating of “feeling” more like a morning or evening person (95). The MEQ consists of nineteen questions regarding preferred timing of daily activities. Output from this questionnaire provides a composite score ranging from 16-86 that can be used to categorize an individual as a definite morning type (70-86), moderate morning type (59- 69), neither type (42-58), moderate evening type (31-41), or definite evening type (16- 30). A reduced version of the MEQ is also available (rMEQ) (114). This version of the chronotype questionnaire was validated using activity rhythms (measured with accelerometers) and showed that evening types had a significantly delayed activity rhythm versus morning and intermediate types (115).

Another questionnaire, the Munich Chronotype Questionnaire (MCTQ), assigns chronotype quantitatively based on the middle of the sleep episode on free days (mid- sleep free days; MSF) (96). Whereas the MEQ assigns chronotype based on preferential timing of daily sleep and activity, the MCTQ measures chronotype based on timing of actual behavior. While the MCTQ does not provide cut-off values to distinguish between early and late chronotypes, comparisons within a sample or population allow for the categorization of “earlier” or “later” chronotypes. The MSF calculated from the MCTQ can also be “corrected” for accumulated sleep debt (MSFsc) (96). Late sleep timing and

19 early alarms during the work week often lead to accumulated sleep debt, and exaggerated and very late free day mid-sleep times. Importantly, MSFsc is unable to be calculated for individuals who have no free days or use alarms on free days (11).

MSFsc on free days can differ by more than 4 hours between earlier and later chronotypes (96). Although variability decreases when comparing sleep/wake timing during the work days, early chronotypes fall asleep almost 2 hours prior to late chronotypes. Interestingly, they still tend to wake up within 30 min of each other, leading to a reduced sleep duration for late chronotypes (96).

Chronotype is known to be affected by environmental and social influences (116). Some evidence also suggests that genetics could contribute to inter-individual variability of chronotype. Duffy et al. (117) demonstrated that individuals who were more of a morning chronotype had a shorter endogenous circadian period, and individuals who were more of an evening chronotype had a longer period. Furthermore, period length was associated with wake time and core body temperature phase. Another study found that a polymorphism in the CLOCK gene was significantly associated with MEQ morningness- eveningness score (118). Chronotype is also significantly associated with dim light melatonin onset, a measure of endogenous phase (98, 99).

Chronotype is not fixed and can vary throughout an individual’s lifespan (119). In a seminal study, Roenneberg and colleagues investigated the impact of age and sex on chronotype of individuals living in Europe. They found that young children and older adults were on average earlier chronotypes, as measured by mid-sleep on free days from the MCTQ, while adolescents and young adults (approximately 20 years of age) had the latest chronotypes (119). Furthermore, women were typically earlier chronotypes than men, until the age of 50 when the gender difference disappeared. A recent study was performed in individuals living in the United States (120). This study used mid-sleep on free (weekend) days as the measure of chronotype. Similarly, this study showed that chronotype continued to become later from age 15 until approximately 20. After 20 years of age chronotype advances steadily throughout the lifespan. Furthermore, a later chronotype was observed in men, however this sex difference was only found until 40 years of age. Following 40 years of age women displayed later chronotypes. In addition,

20 variability in chronotypes (standard deviation in average mid-sleep weekend day) was greatest during the 20-24 age range and decreases non-linearly with age.

1.2.6.2.2 Impact of chronotype on human performance and health

Accumulating evidence has shown that chronotype affects school academic performance (121-123). Specifically, late chronotypes perform more poorly and have lower grades. This is particularly impactful since adolescents have the latest chronotype. The negative association between chronotype and performance has been shown to be strongest in “science-based” academic subjects (chemistry, geography, biology, mathematics) that involve reasoning, logic, and abstract thinking (121). In addition, late chronotypes performed worse on exams administered in the morning compared to morning chronotypes. This difference in performance was abolished when exams were administered in the afternoon (123). Late chronotypes were also sick and absent from school more often than early chrontypes. Thus, chronotype must be taken into account in assessment of academic performance. Some schools have implemented later start times and have observed improvements, primarily in sleepiness and mood (124).

In both adolescents and adults, chronotype is associated with depression (125-127). Specifically, late chronotypes are at a greater risk of having depressive symptoms (125). In contrast one study showed that early chronotypes are more likely to have depression symptoms (128). Late chronotype is also associated with greater likelihood to use addictive substances (113, 125).

Several studies have investigated the relationships between chronotype and dietary choices and obesity (113, 129, 130). Later sleep timing or late chronotype is associated with shorter sleep duration, lower consumption of fruits and vegetables, higher BMI, lower HDL, less frequent meals, and more calories consumed after 8:00 PM (113, 129). After controlling for sleep timing, age, and sleep duration, caloric consumption after 8:00 PM has been shown to be an independent predictor of BMI and total caloric intake. Maukonen et al. (130) investigated the interrelationships between a healthy diet, chronotype, and obesity. Their results showed that evening chronotypes were younger, more likely to smoke, more likely to be physically inactive, more likely to perceive themselves as having poor health, and had higher levels of education.

Merikanto et al. (131) investigated the association between chronotype and cardiovascular disease and type 2 diabetes, as well as outcomes associated with the pathogenesis of these diseases. Results from this cross-sectional study showed that late chronotypes had an increased risk of type 2 diabetes and hypertension compared to morning types. Another study in adolescents showed that late chronotype (measured by MSFsc) is associated with more frequent headaches, stomach aches, and backaches, as well as drinking soft drinks, smoking, and more screen time (132). These results were independent of sleep duration, environmental factors, and concurrent health behaviors. In a recent study, Knutson and Schantz (133) investigated the relationship between chronotype and morbidity and mortality using the UK Biobank data. To determine chronotype, the researchers asked one question to place the subject in a category ranging from definite morning to definite evening. Definite evening types had greater odds of having comorbid conditions, including psychological and neurological disorders, diabetes, renal disease, and CVD.

The impact of chronotype also extends to diurnal variations in physical and cognitive performance. A recent study observed a significant interaction of time of day and chronotype for daytime sleepiness, psychomotor vigilance, executive function, and maximum voluntary contraction (standardized method for measuring muscle strength) (134). In the morning (08:00), early chronotypes performed significantly better than late chronotypes in both cognitive and physical tasks. Unexpectedly, when the data was expressed as a function of time since waking, peak performance on cognitive tasks occurred approximately at the time of typical awaking for early chronotypes and occurred 12.6 hours after waking for late chronotypes. Furthermore, peak performance on a physical task occurred 6.7 hours after waking for early chronotypes and 12.6 hours after waking for late chronotypes. Other literature has shown that chronotype can influence the perceived exertion during exercise sessions, resulting in increased perceived exertion during morning physical activity for late chronotypes (135).

Previous studies have shown that natural light significantly impacts chronotype. Roenneberg and colleagues (96) found that time spent outdoors was significantly correlated with midsleep on free days. People who spent more than 30hours/week

22 outdoors had a midsleep that was almost 2 hours earlier than individuals who spent less than 10 hours/week. In another study, Wright et al. observed melatonin rhythms and sleep timing in 8 participants during 1 week of normal, electric lighting conditions and during 1 week of camping in the natural light/dark environment (136). Following the week of camping, sleep timing advanced, especially in late chronotypes. The authors postulated that natural sunlight synchronized circadian rhythms and decreased circadian disruption. Importantly, during the camping week, physical activity levels were 70% higher, and could have contributed to the advance of the circadian rhythms.

A recent randomized controlled trial recruited late chronotypes for an intervention in which subjects in the experimental condition were asked to participate in practical adjustments to shift their activities to earlier times (e.g. earlier sleep times, morning light exposure, fixed meal times and caffeine intake) (137). The experimental group experienced a 2-hour advance in sleep/wake time and circadian phase (DLMO), accompanied by improvements in morning and afternoon sleepiness, cognition, and physical performance. Most importantly, the experimental group reported significantly reduced depression and stress.

1.2.6.2.3 Social Jetlag

Jetlag is a common form of circadian misalignment in modern society. When individuals travel across time zones, circadian oscillators become desynchronized with the environment. Jetlag caused by trans-meridian travel subsides after a few days, when the phases of circadian oscillators entrain to the new environmental cycle. A phenomenon similar to jetlag, called social jetlag, is misalignment between social obligations and internal circadian rhythms. Social jetlag is chronic circadian misalignment that affects 70% of people living their everyday lives (11). Electrical lighting, TV watching, and smart phone use promote staying up late at night, but work and school schedules require early wake times. Eighty percent of people require an alarm clock to wake up on work days, often several hours before they naturally wake up on the weekend (11). Sleep timing on the weekend (or other days with no social obligations) represents the timing of the internal circadian clock. Thus, social jetlag is quantified as the difference (in hours) between the time of mid-sleep on work days and on free days (138). Similar to

23 chronotype, social jetlag varies with age as well. Social jetlag is greatest in young adults (<21 years old = 3 hours) (96).

Recent studies have found that social jetlag is associated with poor health (11, 13, 55, 139). Roenneberg and colleagues (11) found that social jetlag was positively associated with BMI in overweight individuals. Wong et al. (13) investigated the association between chronotype, social jetlag, and cardiometabolic risk factors in a sample of middle- aged participants. After controlling for sleep quality and various health behaviors, social jetlag associated positively with triglycerides, fasting insulin, insulin resistance, waist circumference, and BMI, and negatively with HDL-cholesterol. Other studies also found that social jetlag was positively associated with BMI, as well as fat mass, cortisol levels, increased heart rate, and the metabolic syndrome (55, 56). In a group of individuals with non-communicable diseases (including type 2 diabetes, hypertension, dyslipidemia, or obesity) social jetlag was found to be significantly positively associated with fasting glucose. In addition, having greater than one 1 hour of social jetlag resulted in twice the likelihood of being overweight (139).

Despite the accumulating evidence of the negative health effects associated with social jetlag, few studies have investigated interventions to reduce this form of circadian misalignment. In a recent study, Zerbini and colleagues (140) developed two different protocols aimed at advancing DLMO and sleep onset, and thereby reducing social jetlag. The two methods the researchers chose were: 1) Wearing blue light-blocking glasses in the evening prior to bed and 2) Opening curtains in the morning to allow increased dawn light exposure. Both methods (imposed on different subjects in separate intervention groups) were designed to expose subjects to light during the advancing portion of the human PRC. Advances in DLMO and work day sleep onset occurred after 1-week wearing light-blocking glasses. However, social jetlag was not significantly improved following either protocol.

1.2.6.2.4 Other measures of circadian misalignment

Other measures to quantify circadian misalignment are also being developed. These measures use analyses of physiological and behavioral rhythms, including DLMO and actigraphy measured sleep data. For example, one method quantifies the relationship

24 between DLMO and sleep onset to determine the circadian alignment between the central circadian rhythm and sleep (Figure 1.5). The average duration between DLMO and sleep onset in young adults is approximately 2 hours. One study found greater insulin resistance when sleep onset was closer to DLMO (e.g., less than 2h) (10).

Two other measures of misalignment that can be performed using only sleep data are composite phase deviation (CPD) and sleep regularity index (SRI) (141, 142). Both of these indices assess the longitudinal regularity of sleep timing. The CPD composite score also takes into account the relationship between sleep timing and the internal circadian rhythm. In a recent study, sleep irregularity was shown to be associated with increased obesity, hypertension, fasting glucose, hemoglobin A1C, stress, and depression (143).

Figure 1.5 Circadian misalignment measured as the phase relationship between DLMO and sleep onset. Young adults have an average phase relationship of 2h between DLMO and sleep onset. Studies have shown that phase relationships less than 2h are associated with poor metabolic health.

1.2.7 Interventions targeting the circadian system

1.2.7.1 Light therapy

Because light is a salient cue for synchronizing the circadian system with the environment, clinical studies have investigated the efficacy of light in shifting the internal circadian rhythm (50, 51). For example, one study found that exposure to the natural outdoor light-dark cycle during a week of camping (e.g. bright sunny days and dark nights) advanced sleep onset, especially in late chronotypes (136). Bright morning light treatment is also used to treat maladies associated with circadian disruption, such as delayed sleep phase disorder and seasonal affective disorder (144-146). While these and other studies largely support the efficacy of light in alleviating circadian misalignment, light treatments have not been widely used, in part because they require special light boxes and optimized exposures to light (e.g., optimal brightness, spectral content, and duration) that have not been standardized (147). In addition, some medications and medical conditions (e.g. glaucoma) can contraindicate safe exposure to light therapy (147) and some subjects report side effects like nausea and headaches (148).

1.2.7.2 Time-restricted feeding

The human circadian system has evolved to promote activity, wakefulness and food intake during the day, and sleep and fasting at night. When food intake occurs during inappropriate times, such as the biological night, it can result in deleterious health consequences, including weight gain (149, 150).

Many studies have investigated the circadian rhythm of food intake in rodents (62, 151). Healthy nocturnal mice eating normal chow show a robust feeding rhythm, with most of their food consumed during the night and very little during the day. When given access to high-fat diet, these mice lose their feeding rhythm and eat during both the day and night. In addition, these mice become obese (152). However, if the same high fat diet is provided only at the proper time of day, health and metabolism improves (62). Hatori et al. (62) showed that when mice were fed high-fat diet only during the night they did not gain excess body weight, in spite of eating the same caloric amount as their ad libitum- fed littermates. They were also protected from hyperinsulinemia, hepatic steatosis, and

26 inflammation. In another study, time restricted (high-fat diet only during the active phase at night) feeding was performed during the 5-day work week and mice ate ad libitum on weekends (151). This scenario, which is relevant to human lifestyles (i.e., less structure on weekends), resulted in the same metabolic benefits and weight improvement as continuous time-restricted feeding.

The timing of food intake has also been explored and implemented as an intervention in humans. A study in college students showed that later timing of food intake, relative to DLMO, was associated with higher body fat percentage and BMI (149). No association was found between caloric content, meal macronutrient content, activity level, or sleep duration and body composition. These findings highlight the importance of the timing of nutritional intake, independent of other factors. Another study tracked daily eating in humans using a smart phone app (153). They found that more than half of the subjects ate for 15 hours or more each day (15h from the first meal to the final meal of the day), which means they were eating during the biological night. The researchers then instructed a group of subjects to restrict their eating to 10-12 hours per day for 16 weeks. At the end of the 16-week intervention, the subjects had an average weight loss of 3.27 kg. Another study found that time of day of eating is associated with total food intake, with increased late-night eating resulting in increased daily caloric consumption (154).

The link between food timing and health outcomes may be explained, in part, by the effects of food timing on circadian clocks in tissues. Food can entrain some circadian oscillators in mammals, independent of the SCN. In rodents, food intake (timing and composition) affects circadian rhythms in peripheral tissues, including the liver (152, 155), but does not alter the SCN rhythm. Very little is known regarding the effect of meal timing on central and peripheral circadian rhythms in humans. One recent study found that delaying daily meals by 5 hours resulted in a similar delay in the phase of the glucose rhythm and a small delay of 1 hour in the rhythm of Per2 mRNA in white adipose tissue (156). The rhythms of melatonin and cortisol were unaltered, suggesting that there was no effect of delayed meal timing on the SCN. These results suggest that, similar to rodents, the timing of food intake can alter peripheral, but not SCN, rhythms in humans.

Based on these findings in mouse models and humans, the timing of food intake is an important factor in managing weight and metabolism.

1.2.7.3 Melatonin

Melatonin is a hormone that is produced rhythmically by the pineal gland. Melatonin has been called the “hormone of darkness” as it is produced primarily during the night. As previously discussed, melatonin is known for its use as a marker of the phase of the endogenous circadian pacemaker. However, melatonin is also able to shift circadian phase (59, 60), perhaps by acting directly on the SCN since melatonin receptors are expressed in the human SCN in tissue collected post-mortem (157). The timing of melatonin administration is critical. The phase response curve for exogenous melatonin is approximately 12 hours out of phase with that of light exposure. Specifically, melatonin in the late afternoon results in phase advances and melatonin in the early morning results in phase delays. Therefore, melatonin should be taken 4-6 hours before bed for the greatest phase advance (158). Oral melatonin administration has no known side effects in adults, other than sleepiness.

1.2.7.4 Physical Activity

The response to regular exercise, including weight loss and training adaptations, exhibits large inter-individual differences, raising the possibility that time of day of exercise may be an important factor (159, 160). Recent studies have observed significantly greater improvements in weight loss, body composition, and energy intake following morning exercise compared to exercise later in the day (160, 161). Specially, one study found that a greater proportion of morning exercisers achieved clinically significant weight loss (>5%) compared to evening exercisers (160). A potential mechanism for the effect of timed exercise could be a phase shift of the human circadian rhythm, which will be discussed in detail later in this review.

1.3 Physical Activity and Human Health

1.3.1 Importance of Physical Activity and Physical Fitness

The prevalence of obesity, cardiovascular disease, and type 2 diabetes is increasing at an alarming rate. Cardiovascular disease is the leading cause of death in the United States and accounts for 1 in 3 deaths (162). An inverse relationship has been shown between physical activity habits and negative health outcomes, including coronary heart disease, cardiovascular disease, stroke, and colon cancer (163). A large cohort study including 39,372 women showed that those who walked as little as 1 hour per week were at a lower risk of coronary heart disease (164). In addition, higher volumes of exercise conferred even lower risk of coronary heart disease, even for women currently at high risk (164). Mortality rates are also 3- to 4-fold greater in low fit compared to high fit individuals (163). The evidence concerning the health benefits of regular exercise also extends to improvements in anxiety, depression, cognition, and overall well-being (1, 2). Researchers have observed these improvements following both resistance and aerobic training (4, 165). For example, Bartholomew and colleagues (4) saw an improvement in positive well-being after one 30-minute bout of moderate intensity aerobic exercise in individuals with major depressive disorder. Larson et al. (166) observed a 32% reduction in the risk for dementia when older individuals exercised 3 or more times each week.

Although the role of physical activity in the prevention of chronic conditions is well- established, nearly half of Americans do not meet the American College of Sports Medicine (ACSM) recommendation of 150 mins of moderate intensity physical exercise combined with muscle strengthening activity 2 days/week (2). In general, few people are participating in physical activity during their leisure time. This change, combined with the decreased amount of time spent in occupations requiring moderate to vigorous physical activities, has contributed to an increase in sedentary lifestyles. Thus, physical inactivity has been considered the biggest health concern of the 21st century (167). Although many large epidemiological studies have shown that physical activity can improve health and reduce the risk of deleterious conditions as well as morbidity and mortality, it is critical to investigate the volume of exercise needed to achieve cardio- protective benefits.

1.3.2 Physical Activity Dose-Response

Consistent participation in physical activity decreases the risk of chronic diseases. To provide useful guidelines to the population at large, it is important to understand the appropriate dose to confer health benefits. The prescribed dose of physical activity is described in terms of frequency, intensity, duration and type of physical activity. Exercise volume is the product of frequency, intensity, and duration, and is often reported in a continuous metric, such as metabolic equivalent (MET). MET is a measure of the energy cost of physical activity. The MET value represents exercise intensity as a multiple of the resting metabolic rate (1 MET = 3.5 mL/kg/min) and is expressed as MET-min or MET- hour per week for physical activity recommendations. The response to physical activity can be measured relative to indicators of health, fitness, or a combination of these factors. Most evidence-based research supports the current ACSM guidelines of 150 minutes per week of moderate physical activity or 75 minutes per week of vigorous physical activity to promote substantial benefits. The evidence for th ese optimal doses is derived from meta-analyses of large epidemiological studies (168, 169).

A number of studies have investigated the dose-response relationship between physical activity and weight management. Brown et al. (170) investigated the determinants of healthy weight maintenance in young Australian women over a 16-year period. They found that women who participated in a volume of exercise equivalent to 500-1000 MET-min/week had higher odds (Odds Ratio: 1.23) of maintaining a healthy BMI at the 16-year follow-up. This volume is equivalent to 150 minutes per week of moderate intensity exercise. In a recent meta-analysis, Liu et al. (168) investigated the dose- response association between physical activity and the incidence of hypertension. In total, this review included 330,222 subjects and found that the risk of hypertension was reduced with increasing levels of physical activity. The individuals that met the minimum physical activity guidelines of 150 min/week (approximately 10 MET hour/week), had a 6% reduced risk of hypertension. Furthermore, the risk of hypertension decreased by 6% more for every 10 MET hour/week increase in physical activity. Huai et al. (171) also pooled data from prospective cohort studies investigating the association between physical activity and incidence of hypertension. They found an inverse dose-response

30 association between recreational physical activity and risk of hypertension. Also, individuals with a greater resting blood pressure typically experienced a greater improvement following an exercise intervention (172).

A higher volume of physical activity also reduces the risk of type 2 diabetes (173, 174). Most meta-analyses report a curvi-linear relationship between volume of physical activity and risk of type 2 diabetes. Thus, the greatest improvement in risk reduction occurred in individuals who were currently sedentary. The greatest risk reduction occurred when individuals progressed from being inactive (0 MET hour/week) to moderately active (6 MET hour/week, RR = 0.77). Therefore, the greatest benefits of increasing physical activity on type 2 diabetes occurs when an individual progresses from sedentary to moderate vigorous activity which meets the ACSM guidelines.

Changes in a health outcome can vary with baseline disease state. Johannsen et al. (175) conducted a randomized trial that consisted of 9 months of training (aerobic, resistance, or combined) to investigate the association between HbA1c improvement and changes in lipid metabolism and skeletal muscle mitochondrial biogenesis. The novel finding from this study was that HbA1C change was independently associated with duration of type 2 diabetes. The authors therefore proposed that physical activity interventions should be performed early after diagnosis to achieve the maximum improvement in glycemic control.

A plethora of literature has shown significant associations between physical activity and cardiovascular health (176, 177). Tanasescu et al. (178) assessed leisure time exercise in a 12-year cohort study involving 44,452 men. Every 2 years a questionnaire was completed by participants that queried time spent in leisure time activities including walking, hiking, jogging, running, heavy outdoor work, or weight training. When the data were analyzed with physical activity as a continuous variable, the results demonstrated that for every 50 MET-hour/week increase in physical activity, the risk ratio of CHD decreased by 26%. In another meta-analysis of 1,683,693 participants, Wahid et al. investigated the association between physical activity and CVD and diabetes (174). The follow-up period was an average of 12.8 years. They found that increasing physical activity by 11.25 MET hour/week was associated with a significant risk reduction of

CVD incidence and CVD mortality (RR = 0.83, 0.77, respectively). Even after adjusting for body weight, these risk ratios were still significantly decreased. The greatest risk reduction occurred when moving from inactive to moderate physical activity (an increase in 6 MET h/week). The authors extrapolated that this risk reduction would equate to a decrease of 4.3% risk of CVD mortality per 1 MET hour/week.

Although evidence for associations between regular physical activity and positive health outcomes is impressive and mounting, there are some discrepancies regarding the manner in which to achieve the appropriate dose. Many studies in recent decades have investigated the effects of fractionation of exercise, or performing exercise in multiple short bouts vs. one long continuous bout. Most evidence-based research suggests that improvements in cardiovascular fitness and overall energy expenditure are similar between short, intermittent daily exercise interventions and long duration, extended interventions (179-181). For example, in one study 47 sedentary women were randomized to a long (30 min) brisk walk, 3 short brisk walks (10 min ea.), or control (no walk) for 10 weeks (181). Maximal oxygen uptake (VO2max) increased similarly between the long and short interval groups, indicating an increase in cardiorespiratory fitness. In another study, 49 sedentary subjects were randomized to a long daily walk, several short daily walks, or a control (no walking prescribed) group for 18 weeks (180). Intensity and total walk duration were held constant for both exercise groups. Reductions in heart rate during a standard submaximal aerobic fitness test (step test) were observed for both walking groups, indicating improved cardiovascular fitness for both groups. Thus, the authors concluded that similar improvements in fitness can be achieved through long duration or accumulated short duration exercise bouts. In contrast, some studies have shown a greater effect of multiple short duration exercise bouts, versus one continuous bout, on daily systolic blood pressure (182, 183). The researchers attributed this finding to an acute exercise training induced reduction in systolic blood pressure referred to as post-exercise hypotension. Thus, interspersed short duration physical activity could be an appropriate therapy to manage hypertension.

One issue to consider concerning the dose-response of physical activity on cardiometabolic health is the considerable inter-individual variability in response to

32 regular physical activity (184). Although evidence for the beneficial effects of regular exercise is undeniable, the issue of inter-individual variability in response to an exercise intervention is often understated. One factor that may contribute to the heterogeneity of exercise response is a genetic factor that predisposes someone as a non-responder. A non- responder can be broadly defined as an individual who experiences no significant change in an outcome of interest. It is possible that non-responders to a particular mode of exercise training may be a responder to another (185, 186).

1.3.3 Physical Activity and Sleep

Epidemiological studies have shown that regular physical activity is associated with improved sleep (187, 188). Importantly, the results from epidemiological studies may not be due to a direct effect of exercise on sleep, but rather an indirect relationship because people who sleep better often have more energy for exercise. In addition, surveys that query an individual’s perception of sleep quality or exercise history may be inaccurate due to issues with recall or truthfulness.

Studies that implemented an exercise intervention found only modest effects on sleep (189, 190). In addition, exercise intervention studies investigated the effects of chronic and acute exercise on subjects with healthy sleep. The limited improvement may have been because the subjects already were healthy sleepers. A meta-analysis by Youngstedt et al. (191) showed that acute exercise had no effect on sleep latency or wake after sleep onset, but a small increase in total sleep duration, slow-wave sleep, and REM sleep latency. In addition, they found that the fitness level did not influence the effect of exercise on sleep. In a study in normal sleepers who recorded physical activity and sleep for 102 days, no association was found between daily exercise duration and sleep variables (189). In this group, when the most active and least active days were compared (without regard to exercise time), a small improvement in wake after sleep onset, and sleep efficiency was observed. However, when time spent outdoors was included as a covariate there was no longer a significant difference. In a separate study involving 7 assessment days the same researchers found (in a between-subjects analysis) only a modest association between mean total energy expenditure and mean subjective sleep latency as well as mean total activity and mean reported insomnia (189). However, no

33 within-subjects associations were found between sleep and daily energy expenditure. Also, no significant differences were found between sleep measures from the most active and least active days. Thus, for each subject, energy expenditure did not affect sleep variables. These two studies conclude that daily physical activity does not promote sleep.

Several studies have investigated the effect of time of day of exercise on sleep parameters with mixed results (72, 192-194). Two studies showed that evening exercise elevated heart rate during sleep (192, 194). One study also found that evening exercise attenuated the normal decrease in core temperature (72) and another found that high intensity exercise near bedtime delayed sleep onset (194). Another study found that exercise before bedtime does not disturb sleep quality (192). Interestingly, one study found that morning exercise improved sleep quality in individuals with insomnia (193). Therefore, overall, evening exercise may negatively impact sleep while morning exercise could improve sleep.

1.3.4 Measurement of Physical Fitness

The American College of Sports Medicine (ACSM) provides categories and working definitions for several components of physical fitness including cardiorespiratory fitness, body composition, muscular strength and endurance, and flexibility (195), which I briefly review below.

1.3.4.1 Cardiorespiratory Fitness

According to ACSM, a maximal graded exercise test (GXT) is a laboratory measure of cardiorespiratory fitness. The goal of a max GXT is to perform incremental exercise up to an intensity that elicits an oxygen uptake that reaches the physiological ceiling of the individual. This physiological ceiling is typically referred to as a VO2max value. The

VO2max represents exercise tolerance as well as cardiovascular oxygen delivery and skeletal muscle oxygen extraction. The criteria used to determine if a VO2max has been -1 achieved is a plateau in oxygen uptake or a change in VO2 ≤ 150 ml* min despite an increase in workload. The VO2max must not be confused with VO2peak. VO2peak is defined as the highest oxygen uptake attained during a testing session when predetermined physiological and subjective criteria have been achieved or exceeded. These criteria

34 include achievement of age-predicted heart rate maximum (Male max HR=220-age; Female max HR=206-0.88*age) (196, 197), a respiratory exchange ratio ≥1.15, a rating of perceived exertion ≥17 (6-20; Borg Scale) (198), and venous blood lactate ≥ 8 mmol. To standardize the max GXT and increase the validity, this set of criteria is used to determine if maximal volitional effort, and thus VO2peak is achieved for each participant. Whether designing standardized or unique GXT protocols for a study design, the overall protocol duration, stage length, and workload increases for each stage should be taken into account (199). It has been suggested that the primary determinant for the differences in VO2max for the general population is differences in stroke volume (200) and/or oxygen extraction by the working skeletal muscle (201, 202).

1.3.4.2 Body Composition

The quantification of components of the human body is important in studying human health. The body can be divided into several organization levels including atomic, molecular, cellular, tissue systems, and whole body (203). Various methods have been used to mathematically model and thus provide estimates of regional and total body composition. Body composition is a useful measure to determine the absolute and relative components of the fat and fat-free masses of an individual.

Anthropometric measures are frequently used in clinical and research settings as a measure of the whole body. These measures include body weight, height, circumferences, and skinfold measures. Body mass index (BMI) is commonly used to measure weight relative to height by the following calculation BMI=kg/m2. BMI is often referenced in clinics and clinical research. Although BMI is unable to distinguish between body fat, muscle mass, or bone, it is an easily obtained measure that is significantly associated with cardiovascular and metabolic outcomes (204).

Currently, dual energy x-ray absorptiometry (DXA) is considered one of the single method criterion measures of total body composition. Based on tissue attenuation of alternating x-ray currents, DXA is capable of estimating the absolute and relative fat mass, bone mineral content mass, and mineral-free lean mass (205). While DXA measurements have error inherit to violations of known assumptions, DXA scans can

35 provide rapid, non-invasive measures of total body and regional measures of body composition using 3-component modeling techniques (206, 207).

1.3.4.3 Muscular Fitness

Muscular strength and endurance are another component of health and fitness (195). Muscular fitness is important because it improves bone health, metabolism, and fat-free mass, and the ability to maintain proper posture and carry out activities of daily living. Tests of muscular fitness are often specific to a certain muscle group. Assessment of total body muscle fitness is not possible. Strength testing assesses the maximal force generated and endurance testing assesses the ability to generate a force for a prolonged period of time.

1.3.4.4 Flexibility

According to ACSM, flexibility is also an important component of physical fitness (195). Flexibility represents the ability to carry out a joint movement through its range of motion. Healthy flexibility allows for proper execution of exercise movements, more easily completing activities of daily living, and a reduction in injury risk (208). Assessments of flexibility are specific to the joint being tested and total body assessments do not exist. The most widely used flexibility assessment is the sit-and-reach test which measures lower back and hamstring flexibility.

1.3.5 Measurement of physical activity

1.3.5.1 Accelerometers

Accelerometers provide objective and reliable measures of physical activity in free-living subjects. Accelerometers are devices that detect human movement in the form of accelerations of motion, often in three planes (tri-axial). The accelerations are converted into activity counts using validated algorithms. Physical activity levels are estimated based on activity counts for a specific time period. Accelerometers also provide estimates of sleep duration and require minimal burden on the subject. These objective measuring devices are often preferred over subjective physical activity questionnaires due to the reduction in the error associated with completion, truthfulness, and recall (209).

Physical activity output measures from accelerometers include time spent in sedentary and moderate to vigorous physical activity, estimations of energy expenditure, and the number of steps taken within specified time intervals. Placement of the accelerometer device depends on the desired measures (210). When estimating physical activity, these devices are commonly worn on a waist belt at the level of the top of the iliac crest, in line with the midpoint of the front of the thigh. For measures of sleep, accelerometers are worn on the wrist where they are more sensitive to movements of the extremities. Accelerometers have clocks that time-stamp activity. This is particularly important when time of day of physical activity and sleep are desired outcome measures.

1.4 Physical Activity and Circadian Rhythms

1.4.1 Phase shifting the circadian clock with exercise

Much is known about the intensity, duration, frequency, volume and progression to maximize the effectiveness of physical activity; however, much less is known concerning the time of day exercise should be completed to maximize the beneficial effects of physical activity.

It is possible that exercise has a time-dependent effect on the circadian system. Proper entrainment of the circadian system to the environment is important for human health, and misalignment of the human circadian system can confer numerous negative health consequences. Exercise, if performed at the appropriate time of day, could shift the phase of the internal circadian rhythm and therefore improve entrainment. In many individuals (e.g., in later chronotypes), an advance of the internal circadian rhythm would better align internal rhythms with the environment and with standard social schedules (Figure 1.6). To date, there is little evidence-based research concerning the optimal timing of exercise. It is known that physical activity at specific times of day alters the phase of the circadian rhythm in rodents (211, 212). Following a single 3-hour exercise pulse, phase advances in the golden hamster were almost 4 hours. In recent years a few studies investigated the effect of exercise on phase shifts of the human circadian system (Table 1.1).

Figure 1.6 Phase advances of the internal circadian rhythm could improve circadian alignment in late chronotypes. The black line is baseline salivary melatonin rhythm for a late chronotype with a DLMO occurring after sleep onset. A phase advance shifts DLMO earlier to before sleep onset.

In one of the earliest studies, Eastman et al. investigated the ability of exercise to improve entrainment for night shift workers by phase delaying the human circadian clock (213). Subjects performed simulated night shift work followed by daytime sleep, thereby delaying sleep by 9 hours. During the night shift, half of the subjects exercised hourly for 15 minutes at 60% maximum heart rate, and the remaining subjects served as controls (no exercise). The rhythm of core body temperature was significantly phase-delayed in the exercise group compared to the controls (6.6 ± 2.5 hrs vs. 4.2 ± 3.4 hrs, respectively). This phase delay meant that the temperature rhythm exhibited a more normal phase relationship with the sleep/wake episode. Furthermore, the subjects in the exercise group reported improved sleep, alertness, and mood.

Table 1.1 Summary of research investigating circadian phase shifts with exercise Time of Day Exercise InterventionA Condition Outcome Activity (Reference) Level

Morning 1 hour, 75% VO2 max, stair Lab 0.5 hour delayB Physically (214) climber, 1 session Active

Morning 1 hour treadmill, 65-75% heart Lab 0.9 hour advance Physically (46) rate reserve Active

Morning 2 hours cycle ergometer 65- Lab 1.2 hour delay (similar to Physically (72) 75% heart rate max control) Active

Morning 30 min, 70% VO2 max, cycle Free living 1 hour advance Physically (215) ergometer, 1 session Active

Mid-day 1 hour, 75% VO2 max, stair Lab 43 minute delayB Physically (214) climber, 1 session Active

Mid-day 1 hour, indoor cardio Free living 1.5 hour advance Not specified (216) equipment, 5 sessions

Mid-day 1 hour treadmill, 65-75% heart Lab 0.7-0.9 hour advance Physically (46) rate reserve Active

Evening 1 hour, 75% VO2 max, stair Lab 0.5 hour advanceB Physically (214) climber, 1 session Active

Evening 30 min, 70% VO2 max, cycle Free-living No Change Physically (215) ergometer, 1 session Active

Evening 1 hour, indoor cardio Free-living 1 hour advance Not specified (216) equipment, 5 sessions

Evening 1 hour treadmill, 65-75% heart Lab 0.5-0.7 hour delay Physically (46) rate reserve Active

Evening 2 hours cycle ergometer 65- Lab 1.3 hour delay (similar to Physically (72) 75% heart rate max control) Active

Evening 45 min, continuous running, 7 Free-living 1.75 hour delay (compared Not specified (89) sessions to morning exercise)

Night 3 hours, alternating arm and leg Lab 1 hour delay Good (217) cycling Physical Condition

Table 1.1 Continued

Night 3 hours, cycle ergometer, 40% Lab Young – 45 min delay No criteria (218) and 60% VO2max, alternating Old – 30 min delay

Night 1 hour, 75% VO2 max, stair Lab 49 minute delayB Physically (214) climber, 1 session Active

Night 1 hour treadmill, 65-75% heart Lab Non-significant Physically (46) rate reserve Active

Night 30 min, 70% VO2 max, cycle Free living 1 hour delay Physically (215) ergometer, 1 session Active

AStudies involving combining exercise with additional intervention were not included. BPost-exercise circadian phase was measured on the same day as the exercise stimulus.

Exercise durations used in several studies were extrapolated from rodent studies and are far too intense for a normal daily exercise routine (e.g. 3 hours/day) (217). Buxton et al. investigated the effect of a single nocturnal session of exercise at one of two durations and intensities: 1 hour of exercise at high intensity or 3 hours at moderate intensity (219).

High intensity exercise was performed on a stair climber at 75% VO2 max and low intensity was maintained at alternating 40% and 60% VO2 max and utilized an arm and leg exerciser. All subjects (n=8 males) performed both conditions as well as a sedentary condition in a within-subjects design. The study protocol consisted of a baseline measure of circadian phase (dim light melatonin onset; DLMO), followed by an exercise condition, and then a post-exercise circadian phase measure, and a 2-week washout period. All study procedures were conducted in constant routine conditions, at a light intensity of 70-80 lux (when not sleeping). Each exercise stimulus occurred at approximately 01:00 hours. This study showed that delays in melatonin onset were similar for both exercise durations and intensities (approximately 1-hour delay for each condition), indicating that more practical exercise recommendations could be used to phase shift the human circadian clock. Since then, other studies have also shown that an exercise bout of 1 hour, or even 30 mins, is capable of shifting circadian rhythms (46, 215).

Time of day of an exercise stimulus can determine whether a phase advance, phase delay, or no shift occurs. Although there is a greater consensus in the literature demonstrating

40 circadian phase delays from exercise, a few studies have also found phase advances (46, 214, 215). Miyazaki et al. performed the first study showing that physical exercise phase advances the central circadian rhythm (61). The researchers conducted a controlled laboratory study involving a shortened sleep-wake schedule (23hr 40min) for 12 cycles in an isolation facility. Each night subjects began sleeping 20 minutes earlier than the previous day. Subjects in the exercise group performed physical activity in the morning and afternoon at an intensity standardized at 140 beats/min. The control group did not perform exercise and were instructed to sit quietly during the same period of time. A phase advance of plasma melatonin occurred in the exercise group while a phase delay was observed in the control group (1.60 ± 0.42 hr and -0.80 ± 0.71 hr, respectively). This study demonstrated that combined morning and afternoon exercise can phase advance the circadian system of humans, which could be used to ameliorate circadian disruption or disorders such as delayed sleep-phase syndrome.

Although Miyazaki and colleagues observed a phase advance following morning and afternoon exercise, other researchers have observed conflicting results. For example, Yamanaka et al. (72) found a similar phase delay was observed following both morning and evening exercise, as well the control group, indicating no effect of time of exercise on the central circadian rhythm. Buxton et al. (214) provided evidence of a phase advancing effect of evening exercise. These researchers investigated the impact of exercise in the morning, afternoon, evening and night on melatonin onset, and observed a sharp break point around the time of melatonin onset (22:31) where phase advances changed to phase delays. The evening exercise resulted in a phase advance on the day of exercise, but this advance was transient, as a large phase delay in melatonin onset was observed the following day. The researchers postulated that the single exercise stimulus was not sufficient to induce a steady-state phase shift in the human circadian system. The authors suggested that repeated exercise exposure at the appropriate circadian phase may be critical to observe a sustained phase shift.

In rodent models, light and exercise are antagonistic when applied within a few hours of each other (220). Youngstedt et al. (220) compared the independent phase-delaying effects of an exercise stimulus with that of bright light exposure during the same time in

41 the night in humans. They also compared these results with that of light exposure and exercise that was separated by a few hours, and occurred at a time when a delay was expected from each stimuli. During each condition, the subjects were in the laboratory for 2.5 days and participated in a sleep/wake protocol consisting of repeated 3 hour cycles of 1 hour of sleep and 2 hours of wake. They collected urine every 90 minutes to measure 6- sulphatoxymelatonin, the urinary metabolite of melatonin. They found a significantly increased phase delay following exercise combined with light treatment (80.8 ± 11.6 min) compared to exercise alone (47.3 ± 21.6 min), and a non-significant increase compared to bright light treatment alone (56.6 ± 15.2 min). The exercise alone group exhibited a phase shift that was 84% of that of the light treatment group. Although previous literature suggested that light is a far greater modulator of the circadian clock, this study showed that properly timed exercise may be an effective alternative.

The majority of previous research on exercise timing in human subjects was conducted in constant routine conditions. Edwards et al. (215) investigated the effect of an exercise stimulus in free-living subjects. The subjects performed one exercise bout on a cycle ergometer for 30 minutes at 70% VO2max at 4 different times throughout the day. Core body temperature data was collected continuously for 24 hours the day before and the day after the single bout of exercise. However, during the exercise day the subjects were able to live freely (unlike previous studies). When the exercise was performed between 4 hours before to 1 hour after the temperature minimum (middle of the night), the subjects experienced a phase delay of the core temperature rhythm (approximately -1 hour). When the exercise was performed between 3 hours to 8 hours after the temperature minimum (morning/early afternoon), the subjects experienced a phase advance in the core temperature rhythm (approximately +1 hour). Exercise performed outside of these times had no change on the phase of the temperature rhythm. Another study investigated the effect of morning and evening exercise in free-living conditions (89). This study involved 7 days of exercise (continuous running) at either 9:00 AM or 9:00 PM. However, exercise intensity was not standardized. The circadian rhythm was assessed by wrist temperature. Evening exercise resulted in a 2-hour phase delay of the wrist temperature rhythm and a slower elevation of skin temperature at the beginning of the sleep episode.

A recent study constructed a human phase response curve to exercise using exercise stimuli at eight different time points throughout the day and night (221). Younstedt et al. (221) found a phase advance following morning and afternoon exercise and a phase delay following evening exercise. This was a rigorous, exhausting protocol that lasted 5.5 days in controlled lab conditions and consisted of 60-minute wake bouts and 30-minute sleep bouts. While the study was rigorous, it does not inform exercise timing recommendations for free-living people.

One difficulty with studying the effects of exercise on circadian rhythms is that exercise induces acute physiological changes that mask the circadian effect. For example, thermoregulatory responses to exercise can acutely affect measurements of core body temperature. There is mixed evidence in the literature concerning the acute effect of exercise on melatonin secretion. Studies have reported increases or decreases in melatonin following evening exercise (222, 223). To avoid masking, it is important to collect the post-exercise circadian phase at least one day after the final exercise stimulus.

Physical activity in general is associated with greater robustness in circadian rhythms. Tranel et al. showed that total daily physical activity was associated with increased amplitude of the wrist temperature rhythm (224). In addition, the amplitude of the temperature rhythm was associated with adiposity and risk factors of the metabolic syndrome. Specifically, lean individuals had a higher temperature rhythm amplitude while obese and metabolically dysfunctional individuals had a low amplitude rhythm. Other studies have shown that low-amplitude core temperature rhythms are associated with neuronal (Parkinson’s, Alzheimer’s), mental (depression), and physiological (metabolic syndrome, obesity) disorders (224-226).

In summary, the majority of reported evidence indicates that exercise can shift circadian rhythms, and it appears that daytime physical activity has the greatest tendency to advance these rhythms (Table 1). However, most of these prior studies were performed in controlled lab conditions and included physically-active adults. It is possible that prior exposure to physical activity could affect the phase shifting response. Investigations on the effects of exercise on circadian parameters in less active individuals is warranted, given the fact that 46% of the U.S. population do not meet the minimum recommended

43 guidelines for aerobic physical activity (227). These previous studies should be used as a guideline for conducting studies in free-living conditions, where subjects are also exposed to other zeitgebers.

1.4.2 Mechanisms causing exercise phase shifts

The mechanisms responsible for exercise-induced phase shifts of the human circadian system are poorly understood. A study in mice indicated that neuropeptide Y and serotonin were both necessary for entrainment to forced treadmill running (228). It has been suggested that exercise-induced phase shifts could be due to an altered retinal sensitivity to light and thus be an effect of concurrent light exposure (219). To explore this phenomenon, Barger el al. (63) tested the effectiveness of moderate intensity exercise to phase delay the circadian system in near constant dark conditions (<5 lux). Eighteen healthy males completed a 49-hour constant routine protocol. Following the constant routine, the subjects underwent a 9-hour delayed sleep-wake cycle. Subjects were randomized to a control or exercise group for 7 days. The exercise group performed three 45-minute bouts each night on a cycle ergometer at 65-75% of heart rate max. The exercise occurred at a time that corresponded to the biological night. When comparing the baseline to post measures of circadian phase, all measures of the melatonin profile in the exercise group were phase delayed by more than 3 hours. The authors noted that the delay experienced in the control (no-exercise) group (approximately 1.5 hours) over the course of the seven experimental days is consistent with what would be expected based upon the intrinsic period of the human circadian system. Results of this study indicate that exercise-induced phase shifts of the central circadian rhythm are independent from photic mechanisms.

1.4.3 Exercise Timing and Performance

Accumulating research shows that exercise performance or capacity fluctuates with a 24- hour rhythm. Skeletal muscles in rodents have self-sustained oscillators and their phases are shifted by timed exercise (229). Likewise, a few studies in humans have shown a 24- hour rhythm of skeletal muscle gene expression and oxidative capacity (230, 231). Thus, it is hypothesized that the skeletal muscle circadian clock influences daily variations in exercise capacity. Anaerobic power performance has been shown to be greatest in the

44 afternoon or early evening. The increase observed has been as high as 7-9% greater during the afternoon or evening, compared to morning (232, 233). One study also found the change in blood glucose following morning exercise was 5-fold higher than that after evening exercise (232).

The master clock in the SCN controls the circadian rhythm of core body temperature, with increases of nearly 1 degree Celsius from the nighttime low to the late afternoon (36.5-37.5°C) (19, 84). Studies have shown that skeletal muscle temperature also increases at this time. Elevated skeletal muscle temperature has been shown to increase power output through a variety of mechanisms. One study found that passive heating of the skeletal muscle increased cycling power output, likely through an increase in muscle fiber conduction velocity, ATP turnover for phosphocreatine utilization, glycolysis, and total anaerobic ATP turnover (234). Resistance exercise performance also displays 24- hour rhythms, with force being greatest in the late afternoon and evening and lowest in the morning (235). In one study, Kuusmaa et al. required subjects to perform combined strength and endurance training in the morning or evening for 24 weeks (236). Following training, isometric strength significantly improved in the evening group, but not in the morning. Furthermore, cardiorespiratory fitness (VO2max) increased similarly for both morning and evening exercise groups. Thus, specific training adaptations may vary based on the timing of the exercise stimuli.

Two recent papers showed a significant effect of time of day of exercise performance and gene expression in mice (237, 238). Greater performance in high intensity exercise occurred in the early active period. Alternatively, greater performance in low and moderate intensity exercise occurred in the late active period. However, this time of day difference was eliminated when the circadian clock was disrupted (Per1 and 2 knockout mice). The results of this study showed that the circadian clock regulates the time-of-day effect on performance. Skeletal muscle gene expression of 513 transcripts were also different in the morning and evening. The authors hypothesized that this could drive the time of day differences in exercise performance. The authors concluded that exercise alters gene transcription in a time-dependent manner, with many of the changes being transcripts involved in metabolic pathways, such as glucose metabolism. The researchers

45 also investigated the effect of a submaximal exercise bout in humans performed at

8:00am and 6:00pm. They found that at the same workload VO2 was significantly reduced in the evening exercise, suggesting improved exercise efficiency.

1.5 Conclusion

Modern lifestyles are characterized by limited exposure to the natural light/dark cycle, activity long after sunset, late night eating, and early morning arousal with the use of alarm clocks. This has resulted in a society that is living in a chronic state of circadian disruption. Numerous studies have shown that circadian disruption is deleterious to health and well-being. Because circadian disruption is common and is associated with poor health, therapeutics that alleviate this disruption could broadly impact human health.

Recent studies have begun to identify an effect of time of day of exercise on human health. However, the topic is understudied and the mechanisms are unknown. Research has shown that exercise can shift circadian rhythms in humans. However, most of these prior studies were performed in controlled lab conditions and used exceptionally long exercise durations, limiting the translation to free-living individuals. Furthermore, the previous studies included physically-active adults so the results may have been confounded by recent non-laboratory exercise regimens. Therefore, the purpose of this dissertation was to investigate the impact of timed exercise on circadian rhythms in free- living, sedentary young adults.

Chapter 2. Circadian rhythm phase shifts caused by timed exercise vary with chronotype in young adults

2.1 Abstract

The human circadian system synchronizes with, or entrains to, the light/dark cycle (sunrise/sunset) to promote activity and food consumption during the day and rest at night. However, strict work schedules and nighttime light exposure impair proper entrainment of the circadian system, resulting in chronic circadian misalignment. Numerous studies have shown that chronic circadian misalignment results in poor health. Although light is the most salient time cue for the circadian system, several laboratory studies have shown that exercise can also entrain the internal circadian rhythm. However, these studies were performed in controlled laboratory conditions with physically-active participants. The purpose of this study was to determine whether timed exercise can phase advance (shift earlier) the internal circadian rhythm in sedentary subjects in free- living conditions. Fifty-two young, sedentary adults (16 male; 24.3±0.76 yr) participated in the study. As a marker of the phase of the internal circadian rhythm, we measured salivary melatonin levels (dim light melatonin onset: DLMO) before and after five days of timed exercise. Participants were randomized to perform either morning (10h after DLMO) or evening (20h after DLMO) supervised exercise training for five consecutive days. Analysis of covariance (ANCOVA) was used to analyze the effect of exercise time on phase shift while adjusting for covariates. We found that morning exercisers had a significantly greater phase advance than evening exercisers. Importantly, the morning exercisers had a 0.6±0.2h phase advance, which could theoretically better align their internal circadian rhythms with the light-dark cycle and with early-morning social obligations. In addition, we found that baseline DLMO, a proxy for chronotype, influenced the effect of timed exercise. We found that for later chronotypes (i.e. subjects with later DLMO), both morning (0.5±0.3h) and evening (0.5±0.2h) exercise advanced the internal circadian rhythm. In contrast, earlier chronotypes (i.e. subjects with earlier DLMO) had phase advances when they exercised in the morning (0.5±0.2h), but phase delays when they exercised in the evening (-0.4±0.3h). Thus, late chronotypes, who experience the most severe circadian misalignment, may benefit from exercise in the

47 morning or evening, but evening exercise may exacerbate circadian misalignment in early chronotypes. Collectively, these results suggest that personalized exercise timing prescription based on chronotype could alleviate circadian misalignment in young adults.

2.2 Introduction

Modern lifestyles can disrupt circadian rhythms. The circadian system controls 24-hour cycles of behavior and physiology, such as rest-activity and feeding rhythms. This system evolved to entrain, or synchronize, to natural cycles of light (sunrise) and dark (sunset) (9). This synchronization between internal and external rhythms has been shown to improve fitness in model organisms (239-241). However, in a modern society, exposure to the natural light/dark cycle is limited. People are active long after sunset (due to electrical lighting), eat late into the night, and wake-up to alarm clocks hours before they would naturally (11, 136). This has resulted in a society that is living in a chronic state of circadian disruption.

Numerous studies have shown that circadian disruption is associated with poor health (11, 57, 100, 106). Shift work, which chronically disrupts the circadian system, increases the risk of cancer, obesity and cardiovascular disease (10-15). Non-shift workers also experience chronic forms of circadian disruption. Seventy percent of the population experiences social jetlag, which is a misalignment between social obligations and internal circadian rhythms (e.g., waking-up to an early-morning alarm clock) (11). Social jetlag is associated with increased cortisol levels and resting heart rate, insulin resistance, obesity, metabolic syndrome, and systemic inflammation (11, 13, 55-57). Late chronotypes, or night owls, often experience social jetlag because they have early-morning work schedules that do not match their natural internal rhythms. Late chronotypes are less physically active, have a higher BMI, and are at increased risk of developing type 2 diabetes and metabolic syndrome (125, 130, 242, 243). Misalignment of circadian rhythms may be a contributing factor leading to obesity and metabolic dysfunction in late chronotypes (244).

Because circadian disruption is common and is associated with poor health, therapeutics that alleviate this disruption could broadly impact human health. The circadian system is remarkably sensitive to environmental cues that synchronize internal rhythms to external cycles. Therefore, behavioral, non-invasive therapeutics could tap into these input circuits and align internal and external rhythms in individuals with disrupted circadian systems. Sunlight is the dominant environmental signal in mammals (51, 245, 246). Studies in

49 rodents have shown that melanopsin-containing retinal ganglion cells in the eye detect irradiance in the environment and send light information directly to the master circadian clock in the suprachiasmatic nucleus (SCN) (31, 247). In rodents, the SCN controls daily rhythms of rest-activity, feeding, body temperature, and hormones (e.g., corticosterone, glucose, insulin, leptin) (29, 30, 83, 248-250). In addition to light, exercise can entrain the circadian system in mammals (61, 63, 211, 220, 251). It is well-known that regular physical activity has substantial health benefits, including improvements in cardiometabolic health, obesity, anxiety, depression, cognition, and overall well-being (1- 6). However, the role of exercise in regulating the circadian system has been understudied. Exercise, if performed at the appropriate time of day, could shift the phase, or timing, of the internal circadian rhythm and therefore improve circadian misalignment. In many individuals (e.g., in later chronotypes), an advance of the internal circadian rhythm would better align internal rhythms with the environment and with standard social schedules. Recent studies in mice showed that exercise performance, gene transcription, and energy utilization depend on the time of day of exercise (237, 238). Several clinical studies showed that morning or early afternoon exercise advances the internal circadian rhythm (46, 61, 215) (but one study showed delays) (72). Other research demonstrated that evening exercise delays the internal circadian rhythm (46, 72) (but one study showed advances) (214). However, most of these prior studies were performed in controlled lab conditions and included physically-active adults so the results may have been confounded by recent non-laboratory exercise regimens. Therefore, in this study we investigated the impact of timed exercise on circadian rhythms in free-living, sedentary young adults.

2.3 Results

2.3.1 Study Participants

Young adults were assessed in the present study because they have later sleep timing (chronotype is latest in young adults) (96) and they could therefore have the greatest benefit from timed exercise, if that exercise advanced their circadian rhythms. Importantly, we also studied sedentary participants (≤2h moderate-vigorous structured exercise/week) and they exercised only during the 30-minute scheduled and supervised session. Sixty-seven participants enrolled in the study (Figure 2.1). Eleven participants

50 withdrew due to personal reasons, including unwillingness to exercise at the prescribed times or to abstain from other physical activity. Four participants were excluded from the analytic data set because we could not measure circadian phase (salivary melatonin levels were erratic or did not exceed the threshold).

Figure 2.1 Participant recruitment. Participants enrolled in the study were 18 to 45 years old, healthy, medication-free (except birth control), and reported no psychiatric or sleep conditions (N=67). Participants were randomized to morning or evening exercise and 11 participants subsequently withdrew. Fifty-two participants were included in the analytic data set (after removing 4 participants because phase could not be determined from the melatonin data).

2.3.2 The time of exercise affects the phase of circadian rhythms in young, sedentary adults

We first determined whether timed exercise shifts the phase of the internal circadian rhythm in young sedentary adults under free-living conditions. We chose morning and evening exercise for two reasons. First, morning exercise has been shown, in some studies, to cause significant phase advances, which could alleviate circadian misalignment in late chronotypes and others whose internal rhythms are delayed relative to the environment (46, 61, 215). Second, morning and evening are the most common times when people exercise on weekdays, so exercise at these times could be

51 implemented as feasible behavioral interventions for circadian disruption in free-living individuals (252). Age, anthropometric (body weight, standing height, BMI, and waist, abdominal and hip circumferences), body composition, (body fat %, fat mass, fat-free mass, mineral-free lean mass) and cardiorespiratory fitness (VO2peak) measures did not differ between the group of participants who exercised in the morning compared to the group who exercised in the evening (baseline characteristics in Table 2.1).

Table 2.1. Characteristics of study participants. VariableA Morning Group Evening Group p-valueB n=26 n=26 Age 24.77 ± 1.20 23.73 ± 0.96 0.50 Body Mass (kg) 66.75 ± 2.84 69.09 ± 2.28 0.52 Height (cm) 165.89 ± 1.30 168.07 ± 1.76 0.33 BMI (kg·m2-1) 24.12 ± 0.85 24.56 ± 0.85 0.72 Anthropometric Waist Circumference (cm) 75.33 ± 1.83 77.59 ± 1.99 0.41 Abdominal Circumference (cm) 82.66 ± 2.11 85.55 ± 2.17 0.34 Hip Circumference (cm) 99.68 ± 1.85 101.03 ± 1.71 0.59 Body Composition Body Fat Percentage (%) 31.11 ± 1.63 32.45 ± 1.81 0.58 Fat Mass (kg) 20.99 ± 1.88 22.28 ± 1.74 0.62 Fat-free Mass (kg) 44.56 ± 1.52 44.82 ± 1.42 0.90 Mineral-free lean Mass (kg) 42.17± 1.46 42.45 ± 1.36 0.89 Cardiorespiratory Fitness -1 -1 VO2Peak (mL·kg ·min ) 38.91 ± 1.55 36.47 ± 1.70 0.29 AData are presented as mean ± SE. BBaseline participant characteristics for the morning and evening exercise groups were compared using independent sample t-tests; p < 0.05.

As a marker of the phase of the internal circadian rhythm, we measured salivary melatonin levels. The rhythmic secretion of melatonin by the pineal gland is under the direct control of the master circadian clock in the SCN (253). The rise in melatonin levels before habitual bedtime (e.g. dim light melatonin onset or DLMO) is a well-established, reliable, and widely-used marker of internal circadian phase of human participants (254). Thus, we measured DLMO before and after 5 days of timed exercise to determine the change in phase caused by exercise. We found there was a trend for morning exercisers to

52 have larger phase advances compared to evening exercisers (Table 2.2: Circadian Phase Shift Unadjusted; Table 2.3: Model 1). When we compared baseline circadian phase (baseline DLMO) between groups, there was a trend for morning exercisers to have an earlier circadian phase (e.g., earlier DLMO) than evening exercisers (Table 2.2). Therefore, we adjusted for the difference in baseline DLMO between the groups and found that morning exercisers had a significantly greater phase advance than evening exercisers (Fig. 2.2; p=0.01, Phase Shift DLMOadjusted in Table 2.3: Model 2). Importantly, the morning exercisers had a 0.6h phase advance, which could theoretically better align their internal circadian rhythms with the light-dark cycle and with early- morning social obligations. All remaining models (Table 2.3) provided non-significant results.

We also analyzed sleep using actigraphy. Sleep onset and mid-sleep during the exercise days were significantly earlier in the morning group compared to the evening group, which is consistent with the trending earlier circadian phase and chronotype in the morning group (Table 2.2). Sleep fragmentation index, a measure of sleep quality, and sleep duration did not differ between the exercise groups. Therefore, despite the earlier timing of sleep in the morning exercise group, the quality and duration of sleep were similar between groups.

Circadian misalignment typically occurs when the timing of sleep (which is often dictated by social and work obligations) is not synchronized with the timing of the internal circadian rhythm (96). One established method for measuring circadian alignment in free- living people is to quantify the duration of time, or phase relationship, between DLMO (the marker of the internal rhythm) and sleep onset (10, 69, 244, 255). Consistent with previous studies of young adults that had circadian misalignment that was associated with metabolic risk (10, 244), we found that the phase relationship was ~2h in our participants and it did not differ between morning and evening exercise groups (Table 2.2).

Table 2.2 Descriptive comparison of Circadian and sleep characteristics stratified by/between exercise timing groups. VariableA Morning Group Evening Group p-valueB Circadian Circadian Phase Shift Unadjusted (h) 0.51 ± 0.21 0.09 ± 0.17 0.13 MEQ Baseline Score 51.35 ± 1.79 46.92 ± 1.86 0.09 Baseline DLMO (hh:mm) 21:50 ± 00:16 22:35 ± 00:17 0.07 Phase Relationship (h) 1.95 ± 0.21 2.54 ± 0.29 0.11 Sleep Sleep Onset (hh:mm) 23:50 ± 00:15 25:06 ± 00:15 <0.01 Mid-sleep (hh:mm) 3:20 ± 00:14 4:52 ± 00:13 <0.01 Sleep Duration (h) 7.02 ± 0.20 7.53 ± 0.26 0.13 Sleep Fragmentation Index (%) 24.50 ± 1.22 26.55 ± 1.80 0.35 Environmental Conditions Civil Daylength (h) 12.91 ± 0.29 12.79 ± 0.29 0.77 AData are presented as mean ± SE. BBaseline participant characteristics for the morning and evening exercise groups were compared using an independent sample t-tests.

Table 2.3 Statistical model to assess covariates Variables Included in Evening vs. Standard p-value ModelA Morning Phase Error Shift Difference Model 1 Exercise Group only -0.42 0.27 0.13 Model 2 Exercise Group adjusted for -0.64 0.25 0.01 baseline DLMO Model 3 Exercise Group adjusted for -0.49 0.28 0.08 baseline MEQ score Model 4 Exercise Group adjusted for -0.53 0.36 0.15 midsleep on exercise days AData were analyzed with ANCOVA using phase shift as the dependent variable and exercise group as the independent variable.

Figure 2.2 Morning exercise advances the phase of the internal circadian rhythm. Young sedentary adults exercised for 5 days either in the morning (n=26) or evening (n=26). Phase shift was calculated as the difference in the timing of DLMO before and after 5 days of exercise and was adjusted for the difference in baseline DLMO between the groups. Data points are raw data. Bars are mean±SEM from ANCOVA model including exercise group and DLMO. *p<0.01

2.3.3 Chronotype is associated with phase shifts in evening exercisers

Previous studies have shown that chronotype, or preferred sleep timing (so-called morning larks or night owls), is highly correlated with internal circadian phase. Moreover, chronotype has been linked to changes in sensitivity of the circadian system to light (256), so it may also modulate the circadian response to exercise. For this study we enrolled young sedentary participants without regard to chronotype. As a result, our participants had variable chronotypes that were normally distributed similar to previous studies of chronotype in young adults (120, 257) (F igure 2.6). We next determined whether the magnitude of phase shifts elicited by timed exercise depended on chronotype. We found that there was no significant correlation between DLMO (a marker

55 of internal phase and a proxy for chronotype) and phase shifts in the morning exercise group (Fig. 2.3A), but there was a strong association between chronotype and phase shifts in the evening exercise group (Fig. 2.3B). Moreover, there was a trend for baseline DLMO to be a modifier for the effect of timed exercise on circadian rhythms (p=0.20). Because these analyses suggested that baseline internal phase modifies the relationship between timed exercise and phase shift, we next examined the impact of morning or evening exercise on phase shift separately in earlier and later chronotypes (Table 2.4; Fig. 2.4). We found that earlier chronotypes had phase advances with morning exercise but phase delays with evening exercise. In contrast, later chronotypes had phase advances with either morning or evening exercise.

Figure 2.3 Phase shift is correlated with internal phase in evening exercisers. Baseline DLMO, which is marker of internal phase and a proxy for chronotype was plotted relative to phase shift by morning (A, n=26) or evening (B, n=26) exercise. Data was analyzed by two-tailed Pearson correlation.

Table 2.4 Model of morning and evening exercise on earlier and later chronotypes Baseline Exercise Mean Standard Evening vs. p-value DLMO Group Error Morning Phase Model 5 groupA Shift Difference Earlier Evening -0.41 0.29 -0.90 0.02 Chronotypes Morning 0.49 0.25 Later Evening 0.46 0.25 -0.08 0.83 Chronotypes Morning 0.54 0.29 AData were analyzed with ANCOVA using phase shift as the dependent variable and exercise group and chronotype group (earlier chronotypes <22:12 baseline DLMO; later chronotypes ≥ 22:12 baseline DLMO) as the independent variables

Figure 2.4 The effects of timed exercise on phase shifts depends on chronotype. Chronotype was dichotomized with a median split and the effect of morning (n=26) and evening (n=26) exercise on phase shift was separately analyzed in earlier (n=26) and later chronotypes (n=26). Data points are raw data. Bars are mean±SEM from two-way ANCOVA model with exercise group and dichotomized DLMO. *p<0.05 vs. all other groups.

2.4 Discussion

Exercise has many well-established benefits to physical and mental health (1-6).The intensity, duration, frequency, mode, volume, and progression of exercise to optimize the beneficial effects have been well-characterized (2, 7, 8). However, the proper timing of exercise, and the additional benefit on circadian rhythms has been understudied. Disruption of circadian rhythms (by shift work, social jetlag, early-morning schedules etc.) is associated with metabolic disorders, cardiovascular disease, and cancer (10-15). If exercise could improve circadian disruption, then it may also improve the risk factors associated with this disruption. A tenet of the mammalian circadian system is that it is differentially sensitive to stimuli given at different times of day. Therefore, it is likely that the response of the internal circadian rhythm to exercise depends on the time of day of exercise. This concept has been addressed in a few studies, and while they consistently

57 showed time-of-day effects of exercise on circadian rhythms, the results were not consistent between the studies (i.e., in some studies morning exercise advanced the circadian rhythm and in others it did not) and they were mostly performed in controlled laboratory conditions that are not translatable (46, 61, 72, 214). Thus, the goal of the present study was to rigorously test the impact of timed exercise on circadian rhythms under free-living conditions and using exercise times and intensities that are feasible for translation.

Overall, in the present study morning exercise advanced the phase of the internal circadian rhythm, whereas evening exercise did not in young sedentary adults. Young adults, on average, have the latest chronotypes and therefore the greatest potential of all ages to experience circadian disruption caused by misalignment of their (later) internal circadian rhythms with early morning obligations (96). Therefore, the present study demonstrated that morning exercise has the greatest potential to alleviate circadian misalignment in young adults. With regard to the feasibility of implementing a morning exercise regimen in a young adult who prefers to stay up late and sleep in, it is important to note that morning exercise does not necessarily mean that an individual has to exercise at 6:00am. In the present st udy, exercise time was scheduled relative to each individual’s internal circadian rhythm and “morning” exercise for a night owl may be at 10:00am. Thus, a morning exercise prescription is feasible even for a late-night person.

The mechanisms by which morning exercise advance the internal circadian rhythm are unknown. It is well established that exposure to morning light can phase advance the internal rhythm (48, 258), therefore, the combination of exposure to morning light and morning exercise, could have additive phase-advancing effects on young adults in free- living conditions. Indeed, one previous study performed in controlled laboratory conditions that combined the two stimuli observed an additive phase-shifting effect (220). Thus, the combination of light and exercise may be a very effective treatment strategy to alleviate social jetlag and the associated metabolic derangements.

Although morning exercise was overall the ideal time to induce phase advances, we also found that baseline DLMO, a proxy for chronotype, influenced the effect of timed exercise. We found that for later chronotypes, both morning and evening exercise

58 advanced the internal circadian rhythm. Thus, late chronotypes, who experience the most severe circadian misalignment, may benefit from exercise in the morning or evening. In contrast, earlier chronotypes had phase advances when they exercised in the morning, but phase delays when they exercised in the evening. These results suggest that evening exercise may exacerbate circadian misalignment in early chronotypes. These results suggest the need for personalized exercise timing prescriptions based on chronotype.

Numerous epidemiological studies have also shown that cumulative years of exposure to circadian disruption (e.g., >5 years) markedly increases the risk of heart disease, breast cancer, and type 2 diabetes compared to acute exposure (259-261). Since the negative effects of circadian disruption increases with years of exposure, it is ideal to target young adults, who are particularly susceptible to circadian misalignment, to reduce the cumulative years of exposure. Our study demonstrates that timed exercise is effective in shifting the phase of the internal circadian rhythm and thus has the potential to reduce early life exposure to circadian disruption and thereby reduce cumulative exposure over a lifetime.

The results of our study in young adults under free-living conditions was in agreement with previous studies in laboratory conditions, such that morning/early afternoon exercise similarly advanced the phase of the internal circadian rhythm in and out of the laboratory (46, 61). Our experimental design had several strengths that make our findings particularly applicable to people in “free-living” conditions. First, the intensity was

“standardized” based on GXTmax results (moderate; a brisk walk or light jog) and duration (30 min/day) of our exercise regimen was practical. This is in contrast to several previous studies that did not standardize the exercise intensity relative to cardiorespiratory fitness (61, 89), or used long (3h) exercise bouts or multiple exercise sessions per day (63, 217, 219, 262). We also chose exercise times that are common in daily routines. Morning and evening are the most accessible times for much of the population. Other studies have investigated the effects of exercise at all times throughout the day and even in the middle of the night, which are not practical for translation to free-living people (61, 63, 213-215, 220, 262). Some previous studies also performed post-exercise phase assessments on the same day as the exercise stimulus, making it difficult to differentiate between an acute

59 effect of exercise or a phase shift (214, 217). In our study, we measured the phase of the internal circadian rhythm on the day after the 5th day of exercise, thereby avoiding the acute masking effect of the exercise session. Finally, although we used well-established and reliable measures of circadian rhythms (DLMO, actigraphy) and cardiorespiratory fitness (GXTMax) and normalized exercise intensity and timing across participants to rigorously test the effect of timed exercise on circadian phase, we can also make these predictions and prescriptions using cheaper and easier methods. For example, we can predict internal phase (and therefore prescribe exercise time) from simple questionnaires (e.g., DLMO is highly correlated with preferred sleep timing) (76).

There were some limitations of our study. We did not design the study to collect baseline sleep measures, but it would have been interesting to determine whether exercise timing affected sleep duration or quality. Also, given that we designed our experiment to examine individuals under free-living conditions, we did not control for other stimuli that could affect the internal circadian phase. For example, travel to the lab for morning exercise could have resulted in increased light exposure during this time, facilitating phase advances independent of or in addition to exercise. We also did not control for sleep timing or duration (e.g. individuals were instructed to sleep as they normally would) and previous research has shown that changing sleep time can affect DLMO (76, 263).

Now that we have identified the proper exercise times to advance circadian rhythms, which can be tailored for chronotype, future studies should investigate the effect of timed exercise on circadian misalignment and risk factors for metabolic disorders, heart disease, and cancer. A timed exercise intervention can easily be translated to clinical practice. While it is not feasible to perform DLMOs on every individual (as they are time- consuming and costly), chronotype questionnaires or self-report of preferred sleep timing could be used as a proxy for DLMO when prescribing the appropriate time to exercise (76). Importantly, exercising at a time of day that elicits a phase advance could increase the already substantial benefits of exercise on health.

2.5 Methods

2.5.1 Participants

Data were collected between July 2017 and March 2019 in Lexington, Kentucky (38.0406° N, 84.5037° W) year-round, except during the 4 weeks following each semi- annual daylight saving time transition. Participants were recruited via local advertisements and were first screened by telephone to query sleep habits, medical diagnoses, and medication use. Participants enrolled in the study were 18 to 45 years old, healthy, medication-free (except birth control), and reported no psychiatric or sleep conditions. All participants were sedentary which was defined as ≤2 hours of structured physical activity per week. Participants had not traveled across more than 1 time zone in the 4 weeks prior to beginning the study and reported no night or rotating shift work in the previous year.

2.5.2 Study Protocol

Immediately following consent, participants completed 2 questionnaires regarding preferred daily timing of sleep and activities, or chronotype (Figure 2.5). Each participant completed a Physical Activity Readiness Questionnaire (PAR-Q) and Health History Form to identify existing contraindications to exercise as specified in the American College of Sports Medicine Guidelines for Exercise Testing and Prescription (264). Height and weight measures were also collected during the consent session. Participants next performed a baseline measure of circadian phase called dim light melatonin onset (DLMO) to determine the timing of the internal circadian rhythm. Next, cardiorespiratory fitness was determined by measuring VO2peak from a maximal graded exercise test

(GXTMax). Circumference measures as well as a DXA scan were performed to assess body composition. Participants then performed 5 days of supervised treadmill exercise in the morning or evening. Participant randomization was performed using Statistical Analysis Software (SAS, Version 9.4). Participants maintained a heart rate corresponding to 70% VO2max for 30 continuous minutes during exercise. One day after the final day of exercise, a post-exercise DLMO was performed to assess changes in the internal rhythm resulting from the exercise intervention.

Figure 2.5 Study protocol. During consent, participants completed questionnaires and height and weight measures. One to 9 days later, we measured baseline circadian phase with the Dim Light Melatonin Onset (DLMO) assay. Next, we took anthropometric (Anthro) and body composition (Body comp) measures and performed a maximal graded exercise test (GXTmax) to determine peak oxygen consumption (VO2peak). Participants then performed 5 days of treadmill exercise, where they maintained a heart rate corresponding to 70% VO2peak for 30 continuous mins, in the morning or evening. One day after the final day of exercise, post-exercise circadian phase was assessed by the DLMO assay.

2.5.3 Chronotype Questionnaires

Each participant completed the Morningness-Eveningness Questionnaire (Self- assessment version, MEQ-SA) to determine chronotype (95). We also collected the Munich Chronotype Questionnaire (MCTQ) to estimate habitual bedtime (96). Previous studies have used the MCTQ to calculate mid-sleep on free dayssleep corrected (MSFsc) as a measure of chronotype, however we were unable to use this measure since 42% of our participants used alarms on free days or had irregular work schedules (11).

2.5.4 Anthropometric and Body Composition Measures

Participants wore lightweight clothing containing no metal and removed shoes for anthropometric and body composition measures. Standing height was determined to the nearest 0.1 cm from a wall-fixed meter stick (Pittsburgh®; Pittsburg, PA) with the participants’ hands positioned on the hips during a maximal inhalation. Body mass was determined to the nearest 0.1 kg using a calibrated electric scale (Escali Corp., Burnsville, MN). Circumference measurements (waist, abdominal, and hip) were taken in triplicate using a fiberglass anthropometric tape (Creative Health Products Inc., Plymouth, MN) and the mean measurement was used in accordance with the guidelines established by the Airlie Conference Proceedings (265). Body composition was measured

62 using total body DXA scans performed using a GE Lunar iDXA (Lunar Inc., Madison, WI) bone densitometer. In accordance with University of Kentucky procedures and policy, all women of reproductive status completed a urine pregnancy test immediately prior to DXA scanning. Only women with a negative urine pregnancy test (within established urine specific gravity ranges) were included. A single, trained investigator analyzed all scans using the Lunar software Version 14.10. Absolute and relative fat mass (kg and %), fat-free mass (kg), and mineral-free lean mass (kg) were determined for each participant.

2.5.5 Actigraphy

Wrist activity was monitored using a triaxial actigraphy device (ActiGraph model wGT3X-BT, Pensacola, FL) worn on the non-dominant wrist for the duration of the study. The actigraphy devices were initialized at 30Hz and participants were instructed to wear the devices at all times, except when they came in contact with water (e.g., bathing). The times when the actigraphy watches were removed (e.g., before bathing) were recorded by participants on their sleep logs. Actigraphy data were downloaded in 10 second epochs and analyzed using AcitLife sotware (version 6.13.3). Wear time validation was performed by indicating non-wear times in the software. Following reintegration to 60 second epochs, the Sadeh algorithm (93), combined with sleep logs, was used to assess sleep timing, duration, and sleep fragmentation. Sleep Fragmentation Index (SFI) represents sleep restlessness, taking into account number of awakenings and number of brief (1-min) sleep bouts (266). Sleep data during the exercise days were included in the analytic dataset. Eight participants were excluded because of missing data when they removed their actigraphy devices during sleep. Due to the masking effect of scheduled morning or evening exercise intervention, cosinor analysis was not appropriate for this data set.

2.5.6 Circadian Phase Measure: Dim Light Melatonin Onset (DLMO)

Participants reported to the lab 8 hours prior to habitual bedtime, which was the sleep timing indicated for free or weekend days (for participants that had no free days) on the MCTQ. Participants were instructed to abstain from taking non-steroidal anti- inflammatory drugs (NSAIDs) throughout the study and to refrain from eating bananas,

63 caffeine, chocolate, and drinking alcohol on the day of the assessment because these have been shown to alter melatonin measurements (78, 267-269). Participants were exposed to <10 lux at eye level, and no light-emitting electronic devices were permitted during DLMO testing to avoid light suppression of melatonin production (71). Saliva samples were collected hourly, except between 2 to 4h before habitual bedtime, when samples were collected every 30 mins (because this is when melatonin levels typically rise) (76), and the final sample was collected at or just after habitual bedtime. Participants were allowed to eat snacks (snacks contained no artificial colorants) except during the 30 minutes prior to sample collection. They were instructed to rinse with water 30 minutes and 15 minutes prior to sampling and to remain seated for 15 mins prior to sample collection. Saliva was collected in salivettes (Sardest, Numbrecht, Germany). Melatonin was measured by SolidPhase Inc. using a BUHLMANN Direct Saliva Melatonin Radio Immunoassay test kit (Schonenbuch, Switzerland). Intra-assay variation coefficients for low, medium, and high levels of salivary melatonin are 20.1%, 4.1%, and 4.8%, respectively. The inter-assay variation coefficients for the assay standards including 0.5, 1.5, 5.0, 15.0 and 50.0 pg·mL-1 for our sample were 6.3%, 7.0%, 5.5%, 4.1%, and 4.8%, respectively. DLMO was calculated as the time when melatonin concentration exceeded and remained above 4pg·ml-1 (linear interpolation) (73, 75). DLMOs were performed at baseline and one day after the final day of exercise (Fig. 2.5). Circadian phase relationship was calculated as the duration of time between baseline DLMO and the averaged sleep onset over the 5 exercise days (244).

2.5.7 Maximal Graded Exercise Test

Maximal graded exercise tests (GXTMax) were completed using an indirect calorimetry testing system with an integrated ECG and a treadmill ergometer (SensorMedics Vmax Encore, CareFusion Corporation, San Diego, CA). During the progressive (speed and grade) tests, oxygen consumption (VO2) and cardiovascular responses (12-lead ECG and heartrate; Cardiosoft v6.7; GE Healthcare/CareFusion Corporation, San Diego, CA) were continuously measured and monitored. The GXTMax tests were performed using progressive 2-min workload stages. The initial stage of the test began with a walking speed of 3.2 mph with a 0.4-mph increase with each subsequent stage. The grade began at

0% and increased by 2% at each successive stage. Test termination was based on volitional fatigue or the presence of any absolute or relative contraindications to continuing according to the American College of Sports Medicine guidelines (264). Each -1 -1 participant achieved VO2peak (ml*kg *min ) determined by the participant’s ability to obtain a minimum of two of the following criteria: respiratory exchange ratio ≥ 1.15, estimated maximal heart rate achieved or exceeded, and/or rating of perceived exertion ≥ 17 (196, 197).

2.5.8 Exercise Intervention

Exercise timing was standardized across participants by scheduling exercise relative to baseline DLMO. During the consent session each participant was randomized to either morning (10 hours after DLMO) or evening (20 hours after DLMO) exercise. Participants were expected to arrive and begin exercise within 1 hour of the scheduled time (ranges in local EST: morning: 5:34-10:17; evening: 15:43-21:43). For 8 participants, DLMO data were not available before beginning exercise, so exercise times were estimated using habitual sleep timing indicated on the MCTQ. Participants exercised for 30 minutes for 5 consecutive days at the designated time. Upon arrival to the lab, and immediately post- exercise, blood pressure and heart rate were measured with an automated system (Omron digital blood pressure monitor HEM-907XL; Hoffman Estates, IL). Exercise training was performed on a treadmill and intensity was maintained at a heart rate corresponding to

70% of VO2peak (determined from GXTMax). Participants wore chest heart rate monitors (Polar, Bethpage, NY) during each training session to ensure prescribed intensity was met. Treadmill speed and/or grade was adjusted to achieve target heart rate. All training sessions were monitored and supervised to ensure that proper heart rate intensities were maintained, to ensure safety of the participants, and to provide motivation.

2.5.9 Phase Shifts

One day after the final day of exercise, each participant performed a post-exercise DLMO. Phase shifts were calculated as the difference (in hours) of post-exercise DLMO subtracted from baseline DLMO. As is convention, phase advances were expressed as positive values and delays as negative values.

2.5.10 Statistical Analysis

Data were analyzed using SAS, Version 9.4. Descriptive data are presented as means ± SEM. Baseline participant characteristics for the morning and evening exercise groups were compared using independent sample t-tests. Analysis of covariance (ANCOVA) was used to analyze the effect of exercise time on phase shift while adjusting for the covariates, baseline DLMO, MEQ, and mid-sleep. A test of interaction between exercise time and baseline DLMO was conducted and Pearson correlation analysis was used to assess associations between baseline DLMO and phase shift for each exercise group. To further examine the effect of baseline circadian phase on phase shift in each exercise group, DLMO was dichotomized into earlier and later chronotype based on a median split (earlier chronotypes <22:12 baseline DLMO; later chronotypes ≥ 22:12 baseline DLMO). A two-way ANCOVA was conducted with an interaction term to compare earlier and later chronotypes.

2.5.11 Study Approval

Prior to participation, written consent was obtained for each participant in accordance with the policies and procedures of the University of Kentucky Institutional Review Board.

2.6 Supplemental Information

Table 2.5 Supplemental Table. Baseline characteristics of study participants by sex Morning Exercise Evening Exercise VariableA Male (N=8)B Female (N=18) Male (N=8) Female (N=18) Age (yr) 21.88 ± 1.49 26.06 ± 1.52 23.63 ± 1.70 23.78 ± 1.19 Body Mass (kg) 73.99 ± 3.65 63.53 ± 3.57 73.10 ± 3.81 67.31 ± 2.79 Height (cm) 171.69 ± 2.11 163.32 ± 1.22 175.54 ± 1.88 164.74 ± 1.95

BMI (kg·m-2) 25.05 ± 0.93 23.71 ± 1.17 23.70 ± 1.06 24.94 ± 1.14 Anthropometric Waist Circumference (cm) 80.76 ± 1.63 72.92 ± 2.35 80.33 ± 3.02 76.37 ± 2.55 Abdominal Circumference (cm) 85.26 ± 1.50 81.51 ± 2.96 85.63 ± 3.86 85.52 ± 2.70 Hip Circumference (cm) 99.79 ± 2.52 99.63 ± 2.47 99.30 ± 2.64 101.79 ± 2.20 Body Composition

-1 -1 VO2 Peak (mL·kg ·min ) 46.38 ± 2.1 35.59 ± 1.47 44.30 ± 2.83 33.25 ± 1.55 Body Fat Percentage (%) 27.88 ± 1.63 32.54 ± 2.19 24.90 ± 3.10 35.81 ± 1.74 Fat Mass (kg) 20.12 ± 1.86 21.37 ± 2.62 18.30 ± 3.01 24.05 ± 2.05 Fat Free Mass (kg) 51.41 ± 1.83 41.52 ± 1.59 52.61 ± 1.86 41.36 ± 1.16 Mineral-free Lean Mass (kg) 48.84 ± 1.78 39.21 ± 1.51 50.02 ± 1.75 39.08 ± 1.09 Cardiorespiratory Fitness

-1 -1 VO2 Peak (mL·kg ·min ) 46.38 ± 2.1 35.59 ± 1.47 44.30 ± 2.83 33.25 ± 1.55 Environmental Conditions Civil Daylength (h) 13.13 ± 0.70 12.81 ± 0.30 12.51 ± 0.39 12.91 ± 0.38 Sleep and Circadian Measures Circadian Phase Shift (h) 0.46 ± 0.42 0.53 ± 0.25 0.25 ± 0.29 0.02 ± 0.21 Baseline DLMO (hh:mm) 22:20 ± 00:27 21:38 ± 00:20 22:33 ± 00:39 22:35 ± 00:19 MEQ Baseline Score 52.38 ± 2.78 50.89 ± 2.32 50.75 ± 2.38 45.22 ± 2.40 Circadian Phase Alignment (h)* 1.41 ± 0.16 2.10 ± 0.26 2.2 ± 0.52 2.69 ± 0.36 Mid-Sleep Exercise Days 3:43 ± 00:23 3:14 ± 00:16 4:38 ± 00:20 4:58 ± 00:17 (hh:mm)* Sleep Fragmentation Index* 27.80 ± 1.91 23.58 ± 1.41 25.96 ± 3.47 26.82 ± 2.16 Sleep Onset (hh:mm)* 24:12 ± 00:22 23:43 ± 00:19 24:42 ± 00:19 25:17 ± 00:21 Sleep Duration (h)* 7.02 ± 0.13 7.02 ± 0.26 7.89 ± 0.34 7.37 ± 0.35 AData are presented as mean ± SE. BN is reduced by 7 total participants.

Figure 2.6 Supplemental figure. Chronotype is normally distributed. Young sedentary adults completed a MEQ questionnaire at baseline. Based on the composite score a majority of the sample was classified as neither type.

Figure 2.7 Supplemental Figure. MEQ is associated with sleep fragmentation during exercise intervention. Young sedentary adults exercised for 5 days either in the morning or evening. Sleep fragmentation, a measure of sleep restlessness, was measured during exercise days using wrist actigraphy.

Figure 2.8 Supplemental figure. Baseline DLMO is associated with mid-sleep during exercise intervention. Young sedentary adults exercised for 5 days either in the morning or evening. Sleep was measured using a combination of wrist actigraphy and sleep logs.

Chapter 3. Expanded Methodology and Procedures

3.1 Study Approval

The study protocol was submitted for full medical review and approved by the Institutional Review Board at the University of Kentucky.

3.2 Subject Recruitment

Subjects were recruited for this study from Lexington and surrounding communities via local advertisements (fliers; Figure 3.1), and verbal and email solicitation. Strategies for recruitment were developed in collaboration with the Participant Recruitment Service Core of the University of Kentucky’s Center for Clinical and Translational Science (CCTS). This included UK Center for Clinical and Translational Research wall mounts, advertisements on internet webpages, and promotion via social media, including Facebook and Twitter advertisements. Subjects were also recruited from registry databases, including ResearchMatch.org and CenterWatch.com.

Following email contact from an interested subject, a phone conversation was scheduled to conduct screening, including explaining additional details to each individual and determining study eligibility. Specifically, subjects were asked about their age, habitual sleep timing, diagnosed conditions, prescribed medications and over-the-counter supplementation, current smoking, recent travel, children, shift work experience, and exercise participation. Subjects reported no psychiatric or sleep conditions and were medication-free, except female contraceptives. All subjects met the predetermined sedentary inclusion criteria, ≤ 2 hours of structured moderate or vigorous physical activity each week. Subjects had not traveled across multiple time zones in the 4 weeks prior to beginning the study and reported no night or rotating shift work in the past year. All subjects provided signed informed consent prior to participation. Following written informed consent, each subject completed a Physical Activity Readiness Questionnaire (PAR-Q) and brief medical history to identify any existing contraindications to exercise testing.

Figure 3.1 Recruitment flyer

3.3 Protocol

3.3.1 Anthropometric measures and body composition

During the first testing session all subjects completed anthropometric measures including standing height and body mass measured with minimal clothing and no shoes. Standing height was determined to the nearest 0.1 cm from a wall-fixed meter stick (Pittsburgh®; Pittsburg, PA) with the participants’ hands positioned on the hips during a maximal inhalation. Body mass was determined to the nearest 0.1 kg using a calibrated electric scale (Escali Corp., Burnsville, MN). BMI was calculated by dividing body mass (kg) by height squared (m2). During a subsequent testing session, measures were collected for waist (narrowest part of the torso), abdomen (level of the umbilicus), and hip

71 circumference (maximal circumference of the buttocks) (195). Circumference measurements were taken in triplicate and the mean measurement was used for analyses using a fiberglass anthropometric tape in accordance with the guidelines established by the Airlie Conference Proceedings for each subject (270).

A dual energy X-ray absorptiometry (DXA) scan was performed on each subject to measure body composition. In accordance with University of Kentucky procedures and policy, all women underwent a urine pregnancy test immediately prior to DXA scanning. In additional, urine specific gravity was also measured using a hand-held refractometer (Agtago, USA, Inc. Bellevue, WA) and compared to established ranges to ensure a valid pregnancy test result. Only women with a negative urine pregnancy test were included as subjects. All scans were performed by a single, trained investigator. Subjects were instructed to wear light clothing and remove shoes, eyeglasses, the accelerometer and any items in their pockets. Subjects were positioned on the center of the scanning platform within the designated scanning area. A Velcro strap was positioned around the feet/ankles, enabling the subject to limit extraneous leg movement. DXA fat mass (g), fat-free mass (g), mineral-free lean mass (g), and percentage fat (%Fat) were assessed.

3.3.2 Cardiorespiratory fitness

To measure cardiorespiratory fitness, a maximal graded exercise test (GXTmax) was conducted at the CCTS, within the Functional Assessment and Body Composition Core

(FABCC). The GXT was performed on a treadmill ergometer while oxygen uptake (VO2) was measured using an indirect calorimetry testing system (SensorMedic Vmax Encore,

CareFusion Corporation, San Diego, CA). VO2 was measured breath by breath and averaged over 1-minute intervals. Heart rate was monitored at baseline, throughout the test, and during recovery using an integrated electrocardiogram (12-lead ECG and heartrate; Cardiosoft v6.7; GE Healthcare/CareFusion Corporation, San Diego, CA). In addition, another heart rate monitor (Polar T31 chest transmitter, Bethpage, NY) was worn on the chest as a back-up to the ECG. In rare circumstances ECG electrodes will lose adhesion and detach from the skin due to excessive sweating. Reading from this secondary heart rate monitor was detected via a compatible watch adhered to the treadmill (Polar A3, Bethpage, NY).

Prior to beginning the GXT, baseline heart rate and blood pressure (assessed by manual auscultation using an appropriate sized cuff and hand aneroid an d stethoscope) were measured while standing, and the ECG was examined to rule out any contraindications to exercise testing according to ACSM guidelines. The treadmill test began at a walking pace and the intensity increased in 2-minute stages. Stage 1 began with each subject walking at 3.2 mph and 0% grade. With each subsequent 2-minute stage the speed increased by 0.4mph and the grade increased by 2.0%. In the final minute of each stage, blood pressure and rating of perceived exertion (RPE; 6-20 Borg Scale(198)) were recorded. Heart rate was recorded in the last ten seconds of each stage. The test termination was based on volitional fatigue or the presence of any absolute or relative contraindications to continuing according to the ACSM guidelines (271). Verbal encouragement was given throughout the test. Upon termination of the test, heart rate, blood pressure, and VO2 measures were collected for five minutes during a passive recovery while the subject remained standing on the treadmill. Confirmation of maximal effort during the GXT was confirmed by meeting two of the following criteria: age predicted heart rate max (male: 220-age (197); female: 206-0.88*age (196)), respiratory exchange ratio ≥1.15, and RPE ≥17.

3.3.3 Circadian and sleep measures

3.3.3.1 Dim light melatonin onset

Dim light melatonin onset (DLMO) was measured using an 8-hour assessment ending at a subject’s habitual bedtime, which was the sleep timing indicated for free or weekend days (for subjects that had no free days) on the MCTQ. For example, if a subject indicated that they “actually get ready to fall asleep” at 1:00am and they need 30 minutes to fall asleep, the DLMO was scheduled for 5:30pm – 1:30am. Prior to beginning data collection subjects were provided with a sample collection sheet that displayed saliva collection times and reminders (Figure 3.2). Subjects were instructed (verbally during consent session and via email) to refrain from alcohol, chocolate, bananas, and any form of caffeine on the day of the DLMO. In addition, subjects were instructed to abstain from taking non-steroidal anti-inflammatory drugs (NSAIDs) during the entire study.

The entirety of this assessment occurred in dim light (≤ 10 lux at eye level) in a bathroom. The bathroom utilized a “double door” entrance for this assessment. This allowed the researcher to enter the dimly lit room without exposing extraneous light on the subject. A researcher was positioned at a desk directly in front of the testing room for the duration of the DLMO. Subjects were instructed to not have any light-emitting electronic devices in the room during the DLMO.

To prepare the room for the DLMO a sign was first adhered to the exterior door indicated testing was in progress. Next, antibacterial air freshener was sprayed on all surfaces of the bathroom. Then, a lamp with an 11-watt bulb was placed in the corner of the room. Using a light meter (Ruby Electronics, Saratoga, CA) the distance at which the light intensity drops below 10 lux was designated with brightly colored tape. Subjects were instructed to not bring their eyes closer to the lamp than the designated line (Figure 3.3). A desk and office chair were placed in the bathroom, just outside of the 10 lux barrier. A blanket was supplied to each subject due to the cold temperatures that were typical of the bathroom. On the desk the following was placed: bottled water (for mouth rinse and hydration), snacks (see below for details), a lab timer, labeled saliva tubes, ice, and the sample collection document.

Snacks were available to the subject during the DLMO. In general, these snacks were dull in color, and contained no artificial colorants. The food items were lacking caffeine, chocolate or banana. Food included apple sauce, granola bars, peanut butter crackers, cheese crackers, and pretzels. Subjects were provided bottles of filtered water. Subjects were allowed to eat anytime except the 30 minutes prior to a sample. If they had recently eaten they were instructed to rinse their mouth vigorously with water 30 minutes prior to a sample. Fifteen minutes prior to each sample, subjects were instructed to rinse with water (regardless if eaten or not) and remain seated until collection of the next sample.

The first saliva sample was collected following 1 hour in dim light. Next, a saliva sample was collected each hour with the final sample collected at or just after habitual bedtime. Between 2-4 hours before habitual bedtime, samples were collected at 30 minute intervals to improve data resolution during this time. During this interval, melatonin concentration was most likely to begin increasing. This time frame was chosen because

DLMO occurs approximately 2 hours prior to sleep onset for normally entrained individuals (76).

Figure 3.2 DLMO collection instruction document

Lights On Dim Light

Figure 3.3 DLMO setup for circadian phase measurement.

For the collection of each saliva sample, salivettes (Sardest, Nümbrecht, Germany) were used. At the time of the collection the subject removed the cylindrical cotton swab from the salivette tube and placed it into his/her mouth. Subjects were instructed to “move the cotton swab around the mouth and gently chew to stimulate saliva flow”. After approximately 1 minute the cotton swab was removed, inserted back into the tube, and placed on ice. Every 2-3 hours a researcher collected the samples and provided fresh ice. Samples were centrifuged for 2 minutes at 4°C and 1.0 rcf. The first sample was collected and centrifuged soon after saliva collection to check for appropriate saliva volume (≥0.5mL). If volume was insufficient the subject was asked to keep the salivette in his/her mouth for additional time.

To assess melatonin concentration in saliva, a double-antibody radioimmunoassay (RIA) was performed by Solidphase Inc (Portland, ME) using a BUHLMANN Direct Saliva Melatonin Radio Immunoassay test kit (Schonenbuch, Switzerland). Intra-assay variation coefficients for low, medium, and high levels of salivary melatonin are 20.1%, 4.1%, and 4.8%, respectively. The inter-assay variation coefficients for low, medium, and high levels of salivary melatonin are 16.7%, 6.6%, and 8.4%, respectively. DLMO was calculated as the time when melatonin concentration exceeded (and remained above) 4pg/ml (linear interpolation; Figure 3.4).

Figure 3.4. Dim light melatonin onset calculation. DLMO was calculated as the time when melatonin concentration exceeded (and remained above) 4pg/ml (A,B).

On the day following the last exercise session, a final post-intervention DLMO was performed. This assessment was usually scheduled at the same time as the baseline assessment. Procedures for this assessment were identical to baseline. Female subjects documented phase of the menstrual cycle by indicating beginning and end of most recent period as well as use of prescribed contraceptive.

Phase shifts were calculated as the difference (in hours) when post-exercise DLMO is subtracted from baseline DLMO. If the post-exercise DLMO was later, that is considered a phase delay and recorded as a negative value. If the post-exercise DLMO is earlier, this is considered a phase advance and is recorded as a positive value.

3.3.3.2 Chronotype

Subjects completed the Munich Chonotype Questionnaire (MCTQ) and a self-assessment version of the Horne-Östberg Morningness-Eveningness Questionnaire (MEQ-SA) to determine chronotype(95, 96). The MCTQ queries typical sleep time for work and work- free days. The outcome variable for the MCTQ is mid-sleep on free days. Mid-sleep on free days is “corrected” for accumulated sleep debt, resulting in MSFsc, by using a previously developed equation (119). Importantly, MSFsc cannot be calculated for individuals with irregular weekly schedules and those who use alarms on free days. Thus, individuals must have at least one day per week of unrestricted sleep to determine chronotype. The MEQ consists of nineteen questions that query the preferred timing for sleep, as well as social, academic, and physical activity events. The result of this questionnaire is a composite score (range: 16-86) separating individuals into 5 categories: definite morning type, moderate morning type, neither type, moderate evening type, definite evening type.

3.3.3.3 Actigraphy and sleep logs

For the duration of the study subjects were instructed to complete daily sleep logs that queried sleep and wake times as well as occurrence of naps, whether or not they were sleeping alone, if the TV remained on while asleep, and if each day was a work or a free day (Figure 3.5). Subjects also wore a triaxial accelerometer (WGT3X-BT Actigraph monitor, Pensacola, FL) on their non-dominant wrist for the entire duration of the study. The accelerometer was initialized using Actilife software (v6.13.3). The device was synced to local computer time, the sampling rate was set at 30 Hz, and demographic data was entered. To preserve battery life idle sleep mode was enabled (device enters low power state after experiencing 10 seconds of inactivity). No stop time was enabled and the accelerometer was able to continue collecting for the life of the battery charge (approximately 30 days). Subjects were instructed to wear the monitor at all times, while sleeping and awake, unless they came in contact with water (e.g. bathing). The time when the monitor was removed was recorded on the provided sleep log (Figure 3.5). Since this monitor could measure light exposure, subjects were instructed to keep long- sleeved clothing pulled above the monitor to allow ambient light exposure.

Figure 3.5 Sleep logs

Actigraphy data was downloaded as 10 second epochs. Then, wear-time validation was performed on the data using vector magnitude. Specifically, periods with a zero value of vector magnitude for ≥5 minutes were flagged as non-wear periods. In combination with the non-wear periods indicated on the subject sleep logs, true non-wear periods were flagged on the actigraphy file. Additional, non-documented non-wear periods were observed and documented.

Following wear-time validation, sleep analysis was performed on each actigraphy file. Data was reintegrated into 60 second epochs for this analysis. This study was not designed to measure baseline sleep data. Sleep data during the 5 exercise days were analyzed and used in the analytic data set. The beginning of the sleep episode was first determined by a reduction in intensity and frequency of activity counts. The end of the sleep episode was determined by an increase in intensity and frequency of activity counts. If sleep logs and sleep period were in concordance, sleep log was recorded as

79 beginning/end of sleep period. If not in concordance, activity counts were used to designate sleep episode times.

The Sadeh algorithm was used to determine sleep onset and offset (93). This algorithm was developed using an adult sample with a mean age of 22.6 yr (current sample: 24.3 yr). The Sadeh algorithm analyzes each epoch to determine if it should be designated as sleep or wake (using the five previous and the five future epochs). After completing the sleep analysis an excel file was exported containing sleep data, including sleep onset, sleep efficiency, total sleep time, total time in bed, wake after sleep onset (WASO), number of awakenings, average awakening length, movement index, and sleep fragmentation index.

3.3.4 Exercise Intervention

During the consent session each subject was randomized to an exercise time. Randomization was determined by a randomization generator function from the SAS statistical analysis software (SAS 9.4, Cary, NC). Results from the baseline DLMO were used to time the exercise. Morning exercise was scheduled for 10 hours after DLMO and evening exercise was scheduled for 20 hours after DLMO. Subjects were expected to arrive and begin exercise within 1 hour of scheduled time. Some subjects were outside of this range due to delays in processing baseline DLMO. Exercise time was estimated for eight subjects using habitual sleep timing indicated on the MCTQ. Subjects exercised for 5 consecutive days at the designated time of day. The intent was to provide the exercise stimulus at a time that was relative to the individuals’ endogenous circadian phase.

Daily exercise occurred at the designated time of day in the University of Kentucky Pediatric Exercise Physiology (PEP) Laboratory (Seaton Building). Time of arrival was documented. Researchers measured blood pressure (Omron digital blood pressure monitor HEM-907XL; Hoffman Estates, IL) at resting and immediately post-exercise. Subjects wore a heart rate monitor (Polar T31 chest transmitter, Bethpage, NY) during each training session to ensure prescribed intensity was met. The exercise session began with a 3-5 minute warm-up prior to beginning the 30 minute exercise bout. Target aerobic training (treadmill walking/jogging) intensity was a heart rate corresponding to

70% of peak VO2 (determined from baseline Max GXT). Target heart rate was

80 maintained within ± 5 beats·min-1. Approximately every 5 minutes of training, heart rate was documented to record that proper intensity was achieved. Treadmill intensity (speed and/or grade) was adjusted to achieve target heart rate. All training sessions were supervised to ensure safety of the subjects and to provide motivation.

3.3.5 Statistical Analysis

Data was analyzed using Statistical Analysis Software (SAS, Version 9.4). Descriptive data were presented as mean ± standard error for all study variables. Baseline subject characteristics for the morning and evening exercise groups were compared using independent sample T-tests. Differences in morning and evening baseline DLMO and MEQ were nearly significant and midsleep during exercise days was significantly different, indicating that these variables could be considered as possible confounders. Using analysis of covariance (ANCOVA) we assessed the effect of exercise time on phase shift while adjusting for covariates that could be influencing the outcome (DLMO, MEQ, midsleep on exercise days). A test of interaction between exercise time and baseline DLMO was conducted. A Pearson correlation analysis was used to assess associations between baseline DLMO and phase shift for each exercise group. To further examine the effect of baseline circadian phase on phase shift in each exercise group, DLMO was dichotomized into early and late chronotype based on a median split (earlier chronotypes <22:12 baseline DLMO; later chronotypes ≥ 22:12 baseline DLMO). A traditional 2-way ANOVA was conducted with an interaction term. The level of significance was set at p<0.05 for all analyses.

3.4 Excluded and Withdrawn Subjects

One female subject (04-S, morning group, Figure 3.6A) was excluded from the analytic data set due to post-exercise saliva melatonin data that was exceeding normal physiological values. Even after diluting the samples 1:8 to get them to fall on a readable portion of the standard curve, the first daytime sample was >200 pg/ml. This typically indicates that the subject is taking exogenous melatonin supplementation. Another female subject (26-S, morning group, Figure 3.6B) was excluded due to a baseline melatonin profile that was very obscure, resulting in erratic increases and decreases in melatonin concentration throughout the sampling period. Two other male subjects (61-S, 74-S,

81 evening group, Figure 3.6C, D) were excluded from the analytic data set because baseline melatonin levels did not exceed the threshold and circadian phase was unable to be calculated.

A male subject (16-S, evening group) withdrew due to academic and personal issues. A female subject (23-S, morning group) withdrew from the study due to a change in her schedule that limited her availability. A female subject (24-S, evening group) withdrew due to a lack of willingness to remain compliant. She was not willing to refrain from additional exercise (outside of study protocol). A male subject (30-S, evening group) withdrew due to an unforeseen event in his family. A female subject (31-S, morning group) withdrew due to scheduling conflicts and difficulty with transportation to testing sessions. A female subject (36-S, morning group) withdrew due to a lack of willingness to remain compliant. She had several physical events occurring soon and was not willing to forgo beginning her training regime. A female subject (44-S, evening group) withdrew due to personal matters that occurred that would adversely affect her sleep and her ability to continue participating in the study. A female subject (45-S, morning group) withdrew during her DLMO because she preferred not to continue with the study procedures. A female subject (52-S, morning group) withdrew because of a separate commitment that required her to begin physical training outside of our study protocol. A female subject (57-S, evening group) withdrew because of an inability to attend scheduled exercise due to work conflicts. A female subject (72-S, morning group) withdrew due to time constraints.

Figure 3.6 Excluded subject DLMOs. Circadian phase was unable to be calculated for four subjects due to (A,B) eratic melatonin profiles or (C,D) low melatonin production not exceeding threshold.

References 1. Hu JP, Guo YH, Wang F, Zhao XP, Zhang QH, and Song QH. Exercise improves cognitive function in aging patients. Int J Clin Exp Med. 2014;7(10):3144-9. 2. Garber CE, Blissmer B, Deschenes MR, Franklin BA, Lamonte MJ, Lee IM, Nieman DC, Swain DP, and American College of Sports M. American College of Sports Medicine position stand. Quantity and quality of exercise for developing and maintaining cardiorespiratory, musculoskeletal, and neuromotor fitness in apparently healthy adults: guidance for prescribing exercise. Med Sci Sports Exerc. 2011;43(7):1334-59. 3. Short KR, Vittone JL, Bigelow ML, Proctor DN, Rizza RA, Coenen-Schimke JM, and Nair KS. Impact of aerobic exercise training on age-related changes in insulin sensitivity and muscle oxidative capacity. Diabetes. 2003;52(8):1888-96. 4. Bartholomew JB, Morrison D, and Ciccolo JT. Effects of acute exercise on mood and well-being in patients with major depressive disorder. Med Sci Sports Exerc. 2005;37(12):2032-7. 5. Ruscheweyh R, Willemer C, Kruger K, Duning T, Warnecke T, Sommer J, Volker K, Ho HV, Mooren F, Knecht S, et al. Physical activity and memory functions: an interventional study. Neurobiol Aging. 2011;32(7):1304-19. 6. Hunter GR, Brock DW, Byrne NM, Chandler-Laney PC, Del Corral P, and Gower BA. Exercise training prevents regain of visceral fat for 1 year following weight loss. Obesity (Silver Spring). 2010;18(4):690-5. 7. The recommended quantity and quality of exercise for developing and maintaining cardiorespiratory and muscular fitness in healthy adults. Position stand of the American College of Sports Medicine. Schweiz Z Sportmed. 1993;41(3):127-37. 8. American College of Sports Medicine position statement on the recommended quantity and quality of exercise for developing and maintaining fitness in healthy adults. Med Sci Sports. 1978;10(3):vii-x. 9. Pittendrigh CS. Temporal organization: reflections of a Darwinian clock-watcher. Annu Rev Physiol. 1993;55(16-54. 10. Baron KG, Reid KJ, Wolfe LF, Attarian H, and Zee PC. Phase Relationship between DLMO and Sleep Onset and the Risk of Metabolic Disease among Normal Weight and Overweight/Obese Adults. J Biol Rhythms. 2017:748730417745914. 11. Roenneberg T, Allebrandt KV, Merrow M, and Vetter C. Social jetlag and obesity. Curr Biol. 2012;22(10):939-43. 12. Koo YS, Song JY, Joo EY, Lee HJ, Lee E, Lee SK, and Jung KY. Outdoor artificial light at night, obesity, and sleep health: Cross-sectional analysis in the KoGES study. Chronobiology international. 2016;33(3):301-14. 13. Wong PM, Hasler BP, Kamarck TW, Muldoon MF, and Manuck SB. Social Jetlag, Chronotype, and Cardiometabolic Risk. J Clin Endocrinol Metab. 2015;100(12):4612-20. 14. Rybnikova NA, Haim A, and Portnov BA. Does artificial light-at-night exposure contribute to the worldwide obesity pandemic? Int J Obes (Lond). 2016;40(5):815-23.

15. Vyas MV, Garg AX, Iansavichus AV, Costella J, Donner A, Laugsand LE, Janszky I, Mrkobrada M, Parraga G, and Hackam DG. Shift work and vascular events: systematic review and meta-analysis. BMJ. 2012;345(e4800. 16. Refinetti R. Circadian Physiology. CRC Press; 2016. 17. Mairan J-JDd. Paris, France; 1729:35. 18. Bruce CPaV. Rhythmic and Synthetic Processes in Growth. Princeton University Press; 1957. 19. Kleitman N. Sleep and Wakefulness. Chicago The University of Chicago Press; 1963. 20. Aschoff J, Gerecke U, and Wever R. Desynchronization of human circadian rhythms. Jpn J Physiol. 1967;17(4):450-7. 21. Pittendrigh CS. The Entrainment of Circadian Oscillations by Skeleton Photoperiods. Science. 1964;144(3618):565. 22. Daan S. The Colin S. Pittendrigh Lecture. Colin Pittendrigh, Jurgen Aschoff, and the natural entrainment of circadian systems. J Biol Rhythms. 2000;15(3):195- 207. 23. Pittendrigh CS. On Temperature Independence in the Clock System Controlling Emergence Time in Drosophila. Proc Natl Acad Sci U S A. 1954;40(10):1018-29. 24. Brown FA, Jr., and Webb HM. Temperature relations of an endogenous daily rhythmicity in the fiddler crab, Uca. Physiol Zool. 1948;21(4):371-81. 25. Izumo M, Johnson CH, and Yamazaki S. Circadian gene expression in mammalian fibroblasts revealed by real-time luminescence reporting: temperature compensation and damping. Proc Natl Acad Sci U S A. 2003;100(26):16089-94. 26. Pagani L, Semenova EA, Moriggi E, Revell VL, Hack LM, Lockley SW, Arendt J, Skene DJ, Meier F, Izakovic J, et al. The physiological period length of the human circadian clock in vivo is directly proportional to period in human fibroblasts. PLoS One. 2010;5(10):e13376. 27. Pittendrigh CS, and Minis DH. The Entrainment of Circadian Oscillations by Light and Their Role as Photoperiodic Clocks. The American Naturalist. 1964;98(902):261-94. 28. Fitzgerald K, and Zucker I. Circadian organization of the estrous cycle of the golden hamster. Proc Natl Acad Sci U S A. 1976;73(8):2923-7. 29. Stephan FK, and Zucker I. Circadian rhythms in drinking behavior and locomotor activity of rats are eliminated by hypothalamic lesions. Proc Natl Acad Sci U S A. 1972;69(6):1583-6. 30. Moore RY, and Eichler VB. Loss of a circadian adrenal corticosterone rhythm following suprachiasmatic lesions in the rat. Brain research. 1972;42(1):201-6. 31. Ralph MR, Foster RG, Davis FC, and Menaker M. Transplanted suprachiasmatic nucleus determines circadian period. Science. 1990;247(4945):975-8. 32. Bargiello TA, Jackson FR, and Young MW. Restoration of circadian behavioural rhythms by gene transfer in Drosophila. Nature. 1984;312(5996):752-4. 33. Zehring WA, Wheeler DA, Reddy P, Konopka RJ, Kyriacou CP, Rosbash M, and Hall JC. P-element transformation with period locus DNA restores rhythmicity to mutant, arrhythmic Drosophila melanogaster. Cell. 1984;39(2 Pt 1):369-76.

34. Hardin PE, Hall JC, and Rosbash M. Feedback of the Drosophila period gene product on circadian cycling of its messenger RNA levels. Nature. 1990;343(6258):536-40. 35. Konopka RJ, and Benzer S. Clock mutants of Drosophila melanogaster. Proc Natl Acad Sci U S A. 1971;68(9):2112-6. 36. Vitaterna MH, King DP, Chang AM, Kornhauser JM, Lowrey PL, McDonald JD, Dove WF, Pinto LH, Turek FW, and Takahashi JS. Mutagenesis and mapping of a mouse gene, Clock, essential for circadian behavior. Science. 1994;264(5159):719-25. 37. Hogenesch JB, Gu YZ, Jain S, and Bradfield CA. The basic-helix-loop-helix-PAS orphan MOP3 forms transcriptionally active complexes with circadian and hypoxia factors. Proc Natl Acad Sci U S A. 1998;95(10):5474-9. 38. Kume K, Zylka MJ, Sriram S, Shearman LP, Weaver DR, Jin X, Maywood ES, Hastings MH, and Reppert SM. mCRY1 and mCRY2 are essential components of the negative limb of the circadian clock feedback loop. Cell. 1999;98(2):193-205. 39. Bunger MK, Wilsbacher LD, Moran SM, Clendenin C, Radcliffe LA, Hogenesch JB, Simon MC, Takahashi JS, and Bradfield CA. Mop3 is an essential component of the master circadian pacemaker in mammals. Cell. 2000;103(7):1009-17. 40. Reppert SM, and Weaver DR. Coordination of circadian timing in mammals. Nature. 2002;418(6901):935-41. 41. Brown AJ, Pendergast JS, and Yamazaki S. Peripheral Circadian Oscillators. Yale J Biol Med. 2019;92(2):327-35. 42. Yamazaki S, Numano R, Abe M, Hida A, Takahashi R, Ueda M, Block GD, Sakaki Y, Menaker M, and Tei H. Resetting central and peripheral circadian oscillators in transgenic rats. Science. 2000;288(5466):682-5. 43. Yoo SH, Yamazaki S, Lowrey PL, Shimomura K, Ko CH, Buhr ED, Siepka SM, Hong HK, Oh WJ, Yoo OJ, et al. PERIOD2::LUCIFERASE real-time reporting of circadian dynamics reveals persistent circadian oscillations in mouse peripheral tissues. Proc Natl Acad Sci U S A. 2004;101(15):5339-46. 44. Izumo M, Pejchal M, Schook AC, Lange RP, Walisser JA, Sato TR, Wang X, Bradfield CA, and Takahashi JS. Differential effects of light and feeding on circadian organization of peripheral clocks in a forebrain Bmal1 mutant. Elife. 2014;3( 45. Aschoff J. Exogenous and endogenous components in circadian rhythms. Cold Spring Harb Symp Quant Biol. 1960;25(11-28. 46. Youngstedt SD, Elliott JA, and Kripke DF. Human circadian phase-response curves for exercise. J Physiol. 2019;597(8):2253-68. 47. Johnson CH, Elliott JA, and Foster R. Entrainment of circadian programs. Chronobiol Int. 2003;20(5):741-74. 48. Khalsa SB, Jewett ME, Cajochen C, and Czeisler CA. A phase response curve to single bright light pulses in human subjects. J Physiol. 2003;549(Pt 3):945-52. 49. Minors DS, Waterhouse JM, and Wirz-Justice A. A human phase-response curve to light. Neurosci Lett. 1991;133(1):36-40. 50. Rahman SA, St Hilaire MA, Chang AM, Santhi N, Duffy JF, Kronauer RE, Czeisler CA, Lockley SW, and Klerman EB. Circadian phase resetting by a single short-duration light exposure. JCI Insight. 2017;2(7):e89494.

51. Czeisler CA, Allan JS, Strogatz SH, Ronda JM, Sanchez R, Rios CD, Freitag WO, Richardson GS, and Kronauer RE. Bright light resets the human circadian pacemaker independent of the timing of the sleep-wake cycle. Science. 1986;233(4764):667-71. 52. Johnson CH. Forty years of PRCs--what have we learned? Chronobiol Int. 1999;16(6):711-43. 53. Rosenwasser AM, Boulos Z, and Terman M. Circadian feeding and drinking rhythms in the rat under complete and skeleton photoperiods. Physiol Behav. 1983;30(3):353-9. 54. Czeisler CA, Duffy JF, Shanahan TL, Brown EN, Mitchell JF, Rimmer DW, Ronda JM, Silva EJ, Allan JS, Emens JS, et al. Stability, precision, and near-24- hour period of the human circadian pacemaker. Science. 1999;284(5423):2177- 81. 55. Parsons MJ, Moffitt TE, Gregory AM, Goldman-Mellor S, Nolan PM, Poulton R, and Caspi A. Social jetlag, obesity and metabolic disorder: investigation in a cohort study. Int J Obes (Lond). 2015;39(5):842-8. 56. Rutters F, Lemmens SG, Adam TC, Bremmer MA, Elders PJ, Nijpels G, and Dekker JM. Is social jetlag associated with an adverse endocrine, behavioral, and cardiovascular risk profile? J Biol Rhythms. 2014;29(5):377-83. 57. Koopman ADM, Rauh SP, van 't Riet E, Groeneveld L, van der Heijden AA, Elders PJ, Dekker JM, Nijpels G, Beulens JW, and Rutters F. The Association between Social Jetlag, the Metabolic Syndrome, and Type 2 Diabetes Mellitus in the General Population: The New Hoorn Study. J Biol Rhythms. 2017;32(4):359- 68. 58. Wilson Jt, Reid KJ, Braun RI, Abbott SM, and Zee PC. Habitual light exposure relative to circadian timing in delayed sleep-wake phase disorder. Sleep. 2018;41(11). 59. Lewy AJ, Ahmed S, and Sack RL. Phase shifting the human circadian clock using melatonin. Behav Brain Res. 1996;73(1-2):131-4. 60. Sharkey KM, and Eastman CI. Melatonin phase shifts human circadian rhythms in a placebo-controlled simulated night-work study. Am J Physiol Regul Integr Comp Physiol. 2002;282(2):R454-63. 61. Miyazaki T, Hashimoto S, Masubuchi S, Honma S, and Honma KI. Phase- advance shifts of human circadian pacemaker are accelerated by daytime physical exercise. Am J Physiol Regul Integr Comp Physiol. 2001;281(1):R197-205. 62. Hatori M, Vollmers C, Zarrinpar A, DiTacchio L, Bushong EA, Gill S, Leblanc M, Chaix A, Joens M, Fitzpatrick JA, et al. Time-restricted feeding without reducing caloric intake prevents metabolic diseases in mice fed a high-fat diet. Cell metabolism. 2012;15(6):848-60. 63. Barger LK, Wright KP, Jr., Hughes RJ, and Czeisler CA. Daily exercise facilitates phase delays of circadian melatonin rhythm in very dim light. Am J Physiol Regul Integr Comp Physiol. 2004;286(6):R1077-84. 64. Edgar DM, Dement WC, and Fuller CA. Effect of SCN lesions on sleep in squirrel monkeys: evidence for opponent processes in sleep-wake regulation. J Neurosci. 1993;13(3):1065-79.

65. Lydic R, Schoene WC, Czeisler CA, and Moore-Ede MC. Suprachiasmatic region of the human hypothalamus: homolog to the primate circadian pacemaker? Sleep. 1980;2(3):355-61. 66. Mills JN, Minors DS, and Waterhouse JM. Adaptation to abrupt time shifts of the oscillator(s) controlling human circadian rhythms. J Physiol. 1978;285(455-70. 67. Duffy JF, and Dijk DJ. Getting through to circadian oscillators: why use constant routines? J Biol Rhythms. 2002;17(1):4-13. 68. Lewy AJ, Cutler NL, and Sack RL. The endogenous melatonin profile as a marker for circadian phase position. J Biol Rhythms. 1999;14(3):227-36. 69. Burgess HJ, Savic N, Sletten T, Roach G, Gilbert SS, and Dawson D. The relationship between the dim light melatonin onset and sleep on a regular schedule in young healthy adults. Behav Sleep Med. 2003;1(2):102-14. 70. Molina TA, and Burgess HJ. Calculating the dim light melatonin onset: the impact of threshold and sampling rate. Chronobiol Int. 2011;28(8):714-8. 71. Benloucif S, Burgess HJ, Klerman EB, Lewy AJ, Middleton B, Murphy PJ, Parry BL, and Revell VL. Measuring melatonin in humans. J Clin Sleep Med. 2008;4(1):66-9. 72. Yamanaka Y, Hashimoto S, Takasu NN, Tanahashi Y, Nishide SY, Honma S, and Honma K. Morning and evening physical exercise differentially regulate the autonomic nervous system during nocturnal sleep in humans. Am J Physiol Regul Integr Comp Physiol. 2015;309(9):R1112-21. 73. Saxvig IW, Wilhelmsen-Langeland A, Pallesen S, Vedaa O, Nordhus IH, Sorensen E, and Bjorvatn B. Objective measures of sleep and dim light melatonin onset in adolescents and young adults with delayed sleep phase disorder compared to healthy controls. J Sleep Res. 2013;22(4):365-72. 74. Phillips AJK, Vidafar P, Burns AC, McGlashan EM, Anderson C, Rajaratnam SMW, Lockley SW, and Cain SW. High sensitivity and interindividual variability in the response of the human circadian system to evening light. Proc Natl Acad Sci U S A. 2019;116(24):12019-24. 75. Crowley SJ, Suh C, Molina TA, Fogg LF, Sharkey KM, and Carskadon MA. Estimating the dim light melatonin onset of adolescents within a 6-h sampling window: the impact of sampling rate and threshold method. Sleep Med. 2016;20(59-66. 76. Burgess HJ, and Eastman CI. The dim light melatonin onset following fixed and free sleep schedules. J Sleep Res. 2005;14(3):229-37. 77. Deacon S, and Arendt J. Posture influences melatonin concentrations in plasma and saliva in humans. Neurosci Lett. 1994;167(1-2):191-4. 78. Murphy PJ, Myers BL, and Badia P. Nonsteroidal anti-inflammatory drugs alter body temperature and suppress melatonin in humans. Physiol Behav. 1996;59(1):133-9. 79. Parry BL, Berga SL, Mostofi N, Klauber MR, and Resnick A. Plasma melatonin circadian rhythms during the menstrual cycle and after light therapy in premenstrual dysphoric disorder and normal control subjects. J Biol Rhythms. 1997;12(1):47-64. 80. Shibui K, Uchiyama M, Okawa M, Kudo Y, Kim K, Liu X, Kamei Y, Hayakawa T, Akamatsu T, Ohta K, et al. Diurnal fluctuation of sleep propensity and

hormonal secretion across the menstrual cycle. Biol Psychiatry. 2000;48(11):1062-8. 81. Wright KP, Jr., and Badia P. Effects of menstrual cycle phase and oral contraceptives on alertness, cognitive performance, and circadian rhythms during sleep deprivation. Behav Brain Res. 1999;103(2):185-94. 82. Cagnacci A, Soldani R, Laughlin GA, and Yen SS. Modification of circadian body temperature rhythm during the luteal menstrual phase: role of melatonin. J Appl Physiol (1985). 1996;80(1):25-9. 83. Osborne AR, and Refinetti R. Effects of hypothalamic lesions on the body temperature rhythm of the golden hamster. Neuroreport. 1995;6(16):2187-92. 84. Refinetti R, and Menaker M. The circadian rhythm of body temperature. Physiol Behav. 1992;51(3):613-37. 85. Edwards BR, T. Waterhouse, J. Zeitgeber-effects of exercise on human circadian rhythms: what are alternative approacehes to investigating the existence of a phase-reponse curve to exercise? Biological Rhythm Research. 2009;40(1):53-69. 86. Atkinson G, Edwards B, Reilly T, and Waterhouse J. Exercise as a synchroniser of human circadian rhythms: an update and discussion of the methodological problems. Eur J Appl Physiol. 2007;99(4):331-41. 87. Edwards B, Waterhouse J, Reilly T, and Atkinson G. A comparison of the suitabilities of rectal, gut, and insulated axilla temperatures for measurement of the circadian rhythm of core temperature in field studies. Chronobiol Int. 2002;19(3):579-97. 88. Daanen HA, van Es EM, and de Graaf JL. Heat strain and gross efficiency during endurance exercise after lower, upper, or whole body precooling in the heat. Int J Sports Med. 2006;27(5):379-88. 89. Rubio-Sastre P, Gomez-Abellan P, Martinez-Nicolas A, Ordovas JM, Madrid JA, and Garaulet M. Evening physical activity alters wrist temperature circadian rhythmicity. Chronobiol Int. 2014;31(2):276-82. 90. Hasselberg MJ, McMahon J, and Parker K. The validity, reliability, and utility of the iButton(R) for measurement of body temperature circadian rhythms in sleep/wake research. Sleep Med. 2013;14(1):5-11. 91. Refinetti R. Relationship between the daily rhythms of locomotor activity and body temperature in eight mammalian species. Am J Physiol. 1999;277(5):R1493- 500. 92. Werth E, Savaskan E, Knoblauch V, Gasio PF, van Someren EJ, Hock C, and Wirz-Justice A. Decline in long-term circadian rest-activity cycle organization in a patient with dementia. J Geriatr Psychiatry Neurol. 2002;15(1):55-9. 93. Sadeh A, Sharkey KM, and Carskadon MA. Activity-based sleep-wake identification: an empirical test of methodological issues. Sleep. 1994;17(3):201- 7. 94. Youngstedt SD, Kripke DF, Elliott JA, and Klauber MR. Circadian abnormalities in older adults. J Pineal Res. 2001;31(3):264-72. 95. Horne JA, and Ostberg O. A self-assessment questionnaire to determine morningness-eveningness in human circadian rhythms. Int J Chronobiol. 1976;4(2):97-110.

96. Roenneberg T, Wirz-Justice A, and Merrow M. Life between clocks: daily temporal patterns of human chronotypes. J Biol Rhythms. 2003;18(1):80-90. 97. Smith CS, Reilly C, and Midkiff K. Evaluation of three circadian rhythm questionnaires with suggestions for an improved measure of morningness. J Appl Psychol. 1989;74(5):728-38. 98. Kantermann T, Sung H, and Burgess HJ. Comparing the Morningness- Eveningness Questionnaire and Munich ChronoType Questionnaire to the Dim Light Melatonin Onset. J Biol Rhythms. 2015;30(5):449-53. 99. Kitamura S, Hida A, Aritake S, Higuchi S, Enomoto M, Kato M, Vetter C, Roenneberg T, and Mishima K. Validity of the Japanese version of the Munich ChronoType Questionnaire. Chronobiol Int. 2014;31(7):845-50. 100. Boggild H, and Knutsson A. Shift work, risk factors and cardiovascular disease. Scand J Work Environ Health. 1999;25(2):85-99. 101. Brum MC, Filho FF, Schnorr CC, Bottega GB, and Rodrigues TC. Shift work and its association with metabolic disorders. Diabetol Metab Syndr. 2015;7(45. 102. Esquirol Y, Bongard V, Mabile L, Jonnier B, Soulat JM, and Perret B. Shift work and metabolic syndrome: respective impacts of job strain, physical activity, and dietary rhythms. Chronobiol Int. 2009;26(3):544-59. 103. Schernhammer ES, Laden F, Speizer FE, Willett WC, Hunter DJ, Kawachi I, Fuchs CS, and Colditz GA. Night-shift work and risk of colorectal cancer in the nurses' health study. J Natl Cancer Inst. 2003;95(11):825-8. 104. Megdal SP, Kroenke CH, Laden F, Pukkala E, and Schernhammer ES. Night work and breast cancer risk: a systematic review and meta-analysis. Eur J Cancer. 2005;41(13):2023-32. 105. Ballard T, Lagorio S, De Angelis G, and Verdecchia A. Cancer incidence and mortality among flight personnel: a meta-analysis. Aviat Space Environ Med. 2000;71(3):216-24. 106. Nakamura K, Shimai S, Kikuchi S, Tominaga K, Takahashi H, Tanaka M, Nakano S, Motohashi Y, Nakadaira H, and Yamamoto M. Shift work and risk factors for coronary heart disease in Japanese blue-collar workers: serum lipids and anthropometric characteristics. Occup Med (Lond). 1997;47(3):142-6. 107. Morris CJ, Purvis TE, Hu K, and Scheer FA. Circadian misalignment increases cardiovascular disease risk factors in humans. Proceedings of the National Academy of Sciences of the United States of America. 2016;113(10):E1402-11. 108. Hansen TW, Li Y, Boggia J, Thijs L, Richart T, and Staessen JA. Predictive role of the nighttime blood pressure. Hypertension. 2011;57(1):3-10. 109. Scheer FA, Hilton MF, Mantzoros CS, and Shea SA. Adverse metabolic and cardiovascular consequences of circadian misalignment. Proc Natl Acad Sci U S A. 2009;106(11):4453-8. 110. Buxton OM, Cain SW, O'Connor SP, Porter JH, Duffy JF, Wang W, Czeisler CA, and Shea SA. Adverse metabolic consequences in humans of prolonged sleep restriction combined with circadian disruption. Sci Transl Med. 2012;4(129):129ra43. 111. Vetter C. Circadian disruption: What do we actually mean? Eur J Neurosci. 2018. 112. Roenneberg T, and Merrow M. The Circadian Clock and Human Health. Current biology : CB. 2016;26(10):R432-43.

113. Baron KG, Reid KJ, Kern AS, and Zee PC. Role of sleep timing in caloric intake and BMI. Obesity (Silver Spring). 2011;19(7):1374-81. 114. Adan A, and Almirall H. Horne & Östberg morningness-eveningness questionnaire: A reduced scale. Personality and Individual Differences. 1991;12(3):241-53. 115. Natale V, Esposito MJ, Martoni M, and Fabbri M. Validity of the reduced version of the Morningness–Eveningness Questionnaire. Sleep and Biological Rhythms. 2006;4(72-4. 116. Porcheret K, Wald L, Fritschi L, Gerkema M, Gordijn M, Merrrow M, Rajaratnam SMW, Rock D, Sletten TL, Warman G, et al. Chronotype and environmental light exposure in a student population. Chronobiol Int. 2018;35(10):1365-74. 117. Duffy JF, Rimmer DW, and Czeisler CA. Association of intrinsic circadian period with morningness-eveningness, usual wake time, and circadian phase. Behav Neurosci. 2001;115(4):895-9. 118. Katzenberg D, Young T, Finn L, Lin L, King DP, Takahashi JS, and Mignot E. A CLOCK polymorphism associated with human diurnal preference. Sleep. 1998;21(6):569-76. 119. Roenneberg T, Kuehnle T, Pramstaller PP, Ricken J, Havel M, Guth A, and Merrow M. A marker for the end of adolescence. Curr Biol. 2004;14(24):R1038- 9. 120. Fischer D, Lombardi DA, Marucci-Wellman H, and Roenneberg T. Chronotypes in the US - Influence of age and sex. PLoS One. 2017;12(6):e0178782. 121. Zerbini G, van der Vinne V, Otto LKM, Kantermann T, Krijnen WP, Roenneberg T, and Merrow M. Lower school performance in late chronotypes: underlying factors and mechanisms. Sci Rep. 2017;7(1):4385. 122. Zerbini G, and Merrow M. Time to learn: How chronotype impacts education. Psych J. 2017;6(4):263-76. 123. van der Vinne V, Zerbini G, Siersema A, Pieper A, Merrow M, Hut RA, Roenneberg T, and Kantermann T. Timing of examinations affects school performance differently in early and late chronotypes. J Biol Rhythms. 2015;30(1):53-60. 124. Boergers J, Gable CJ, and Owens JA. Later school start time is associated with improved sleep and daytime functioning in adolescents. J Dev Behav Pediatr. 2014;35(1):11-7. 125. Kivela L, Papadopoulos MR, and Antypa N. Chronotype and Psychiatric Disorders. Curr Sleep Med Rep. 2018;4(2):94-103. 126. Antypa N, Vogelzangs N, Meesters Y, Schoevers R, and Penninx BW. Chronotype Associations with Depression and Anxiety Disorders in a Large Cohort Study. Depress Anxiety. 2016;33(1):75-83. 127. Haraden DA, Mullin BC, and Hankin BL. The relationship between depression and chronotype: A longitudinal assessment during childhood and adolescence. Depress Anxiety. 2017;34(10):967-76. 128. Lemoine P, Zawieja P, and Ohayon MM. Associations between morningness/eveningness and psychopathology: an epidemiological survey in three in-patient psychiatric clinics. J Psychiatr Res. 2013;47(8):1095-8.

129. Lucassen EA, Zhao X, Rother KI, Mattingly MS, Courville AB, de Jonge L, Csako G, Cizza G, and Sleep Extension Study G. Evening chronotype is associated with changes in eating behavior, more sleep apnea, and increased stress hormones in short sleeping obese individuals. PLoS One. 2013;8(3):e56519. 130. Maukonen M, Kanerva N, Partonen T, Kronholm E, Konttinen H, Wennman H, and Mannisto S. The associations between chronotype, a healthy diet and obesity. Chronobiol Int. 2016;33(8):972-81. 131. Merikanto I, Lahti T, Puolijoki H, Vanhala M, Peltonen M, Laatikainen T, Vartiainen E, Salomaa V, Kronholm E, and Partonen T. Associations of chronotype and sleep with cardiovascular diseases and type 2 diabetes. Chronobiol Int. 2013;30(4):470-7. 132. Griepy G DI, Whitehead RD, Elgar FJ. More than just sleeping in: a late timing of sleep is associated with health problems and unhealthy behaviours in adolescents. Sleep Medicine 2018. 133. Knutson KL, and von Schantz M. Associations between chronotype, morbidity and mortality in the UK Biobank cohort. Chronobiol Int. 2018;35(8):1045-53. 134. Facer-Childs ER, Boiling S, and Balanos GM. The effects of time of day and chronotype on cognitive and physical performance in healthy volunteers. Sports Med Open. 2018;4(1):47. 135. Rossi A, Formenti D, Vitale JA, Calogiuri G, and Weydahl A. The Effect of Chronotype on Psychophysiological Responses during Aerobic Self-Paced Exercises. Percept Mot Skills. 2015;121(3):840-55. 136. Wright KP, Jr., McHill AW, Birks BR, Griffin BR, Rusterholz T, and Chinoy ED. Entrainment of the human circadian clock to the natural light-dark cycle. Curr Biol. 2013;23(16):1554-8. 137. Facer-Childs ER, Middleton B, Skene DJ, and Bagshaw AP. Resetting the late timing of 'night owls' has a positive impact on mental health and performance. Sleep Med. 2019. 138. Wittmann M, Dinich J, Merrow M, and Roenneberg T. Social jetlag: misalignment of biological and social time. Chronobiol Int. 2006;23(1-2):497- 509. 139. Mota MC, Silva CM, Balieiro LCT, Fahmy WM, and Crispim CA. Social jetlag and metabolic control in non-communicable chronic diseases: a study addressing different obesity statuses. Sci Rep. 2017;7(1):6358. 140. Zerbini G, Kantermann T, and Merrow M. Strategies to decrease social jetlag: Reducing evening blue light advances sleep and melatonin. Eur J Neurosci. 2018. 141. Fischer D, Vetter C, and Roenneberg T. A novel method to visualise and quantify circadian misalignment. Sci Rep. 2016;6(38601. 142. Phillips AJK, Clerx WM, O'Brien CS, Sano A, Barger LK, Picard RW, Lockley SW, Klerman EB, and Czeisler CA. Irregular sleep/wake patterns are associated with poorer academic performance and delayed circadian and sleep/wake timing. Sci Rep. 2017;7(1):3216. 143. Lunsford-Avery JR, Engelhard MM, Navar AM, and Kollins SH. Validation of the Sleep Regularity Index in Older Adults and Associations with Cardiometabolic Risk. Sci Rep. 2018;8(1):14158.

144. Rosenthal NE, Joseph-Vanderpool JR, Levendosky AA, Johnston SH, Allen R, Kelly KA, Souetre E, Schultz PM, and Starz KE. Phase-shifting effects of bright morning light as treatment for delayed sleep phase syndrome. Sleep. 1990;13(4):354-61. 145. Lam RW, Terman M, and Wirz-Justice A. Light therapy for depressive disorders: indications and efficacy. Mod Probl Pharmacopsychiatry. 1997;25(215-34. 146. Virk G, Reeves G, Rosenthal NE, Sher L, and Postolache TT. Short exposure to light treatment improves depression scores in patients with seasonal affective disorder: A brief report. Int J Disabil Hum Dev. 2009;8(3):283-6. 147. Gooley JJ. Treatment of circadian rhythm sleep disorders with light. Ann Acad Med Singapore. 2008;37(8):669-76. 148. Terman M, and Terman JS. Bright light therapy: side effects and benefits across the symptom spectrum. J Clin Psychiatry. 1999;60(11):799-808; quiz 9. 149. McHill AW, Phillips AJ, Czeisler CA, Keating L, Yee K, Barger LK, Garaulet M, Scheer FA, and Klerman EB. Later circadian timing of food intake is associated with increased body fat. Am J Clin Nutr. 2017;106(5):1213-9. 150. Arble DM, Bass J, Laposky AD, Vitaterna MH, and Turek FW. Circadian timing of food intake contributes to weight gain. Obesity (Silver Spring). 2009;17(11):2100-2. 151. Chaix A, Zarrinpar A, Miu P, and Panda S. Time-restricted feeding is a preventative and therapeutic intervention against diverse nutritional challenges. Cell Metab. 2014;20(6):991-1005. 152. Pendergast JS, Branecky KL, Yang W, Ellacott KL, Niswender KD, and Yamazaki S. High-fat diet acutely affects circadian organisation and eating behavior. The European journal of neuroscience. 2013;37(8):1350-6. 153. Gill S, and Panda S. A Smartphone App Reveals Erratic Diurnal Eating Patterns in Humans that Can Be Modulated for Health Benefits. Cell Metab. 2015;22(5):789-98. 154. de Castro JM. The time of day of food intake influences overall intake in humans. J Nutr. 2004;134(1):104-11. 155. Stokkan KA, Yamazaki S, Tei H, Sakaki Y, and Menaker M. Entrainment of the circadian clock in the liver by feeding. Science. 2001;291(5503):490-3. 156. Wehrens SMT, Christou S, Isherwood C, Middleton B, Gibbs MA, Archer SN, Skene DJ, and Johnston JD. Meal Timing Regulates the Human Circadian System. Curr Biol. 2017;27(12):1768-75 e3. 157. Reppert SM, Weaver DR, Rivkees SA, and Stopa EG. Putative melatonin receptors in a human biological clock. Science. 1988;242(4875):78-81. 158. Lewy AJ. Clinical applications of melatonin in circadian disorders. Dialogues Clin Neurosci. 2003;5(4):399-413. 159. King NA, Hopkins M, Caudwell P, Stubbs RJ, and Blundell JE. Individual variability following 12 weeks of supervised exercise: identification and characterization of compensation for exercise-induced weight loss. Int J Obes (Lond). 2008;32(1):177-84. 160. Willis EA, Creasy SA, Honas JJ, Melanson EL, and Donnelly JE. The effects of exercise session timing on weight loss and components of energy balance: midwest exercise trial 2. Int J Obes (Lond). 2019.

161. Alizadeh Z, Younespour S, Rajabian Tabesh M, and Haghravan S. Comparison between the effect of 6 weeks of morning or evening aerobic exercise on appetite and anthropometric indices: a randomized controlled trial. Clin Obes. 2017;7(3):157-65. 162. Benjamin EJ, Blaha MJ, Chiuve SE, Cushman M, Das SR, Deo R, de Ferranti SD, Floyd J, Fornage M, Gillespie C, et al. Heart Disease and Stroke Statistics-2017 Update: A Report From the American Heart Association. Circulation. 2017;135(10):e146-e603. 163. Blair SN, Cheng Y, and Holder JS. Is physical activity or physical fitness more important in defining health benefits? Med Sci Sports Exerc. 2001;33(6 Suppl):S379-99; discussion S419-20. 164. Lee IM, Rexrode KM, Cook NR, Manson JE, and Buring JE. Physical activity and coronary heart disease in women: is "no pain, no gain" passe? JAMA. 2001;285(11):1447-54. 165. Bibeau WS, Moore JB, Mitchell NG, Vargas-Tonsing T, and Bartholomew JB. Effects of acute resistance training of different intensities and rest periods on anxiety and affect. J Strength Cond Res. 2010;24(8):2184-91. 166. Larson EB, Wang L, Bowen JD, McCormick WC, Teri L, Crane P, and Kukull W. Exercise is associated with reduced risk for incident dementia among persons 65 years of age and older. Ann Intern Med. 2006;144(2):73-81. 167. Blair SN. Physical inactivity: the biggest public health problem of the 21st century. Br J Sports Med. 2009;43(1):1-2. 168. Liu X, Zhang D, Liu Y, Sun X, Han C, Wang B, Ren Y, Zhou J, Zhao Y, Shi Y, et al. Dose-Response Association Between Physical Activity and Incident Hypertension: A Systematic Review and Meta-Analysis of Cohort Studies. Hypertension. 2017;69(5):813-20. 169. Sattelmair J, Pertman J, Ding EL, Kohl HW, 3rd, Haskell W, and Lee IM. Dose response between physical activity and risk of coronary heart disease: a meta- analysis. Circulation. 2011;124(7):789-95. 170. Brown WJ, Kabir E, Clark BK, and Gomersall SR. Maintaining a Healthy BMI: Data From a 16-Year Study of Young Australian Women. Am J Prev Med. 2016;51(6):e165-e78. 171. Huai P, Xun H, Reilly KH, Wang Y, Ma W, and Xi B. Physical activity and risk of hypertension: a meta-analysis of prospective cohort studies. Hypertension. 2013;62(6):1021-6. 172. Cornelissen VA, and Smart NA. Exercise training for blood pressure: a systematic review and meta-analysis. J Am Heart Assoc. 2013;2(1):e004473. 173. Cloostermans L, Wendel-Vos W, Doornbos G, Howard B, Craig CL, Kivimaki M, Tabak AG, Jefferis BJ, Ronkainen K, Brown WJ, et al. Independent and combined effects of physical activity and body mass index on the development of Type 2 Diabetes - a meta-analysis of 9 prospective cohort studies. Int J Behav Nutr Phys Act. 2015;12(147. 174. Wahid A, Manek N, Nichols M, Kelly P, Foster C, Webster P, Kaur A, Friedemann Smith C, Wilkins E, Rayner M, et al. Quantifying the Association Between Physical Activity and Cardiovascular Disease and Diabetes: A Systematic Review and Meta-Analysis. J Am Heart Assoc. 2016;5(9).

175. Johannsen NM, Sparks LM, Zhang Z, Earnest CP, Smith SR, Church TS, and Ravussin E. Determinants of the Changes in Glycemic Control with Exercise Training in Type 2 Diabetes: A Randomized Trial. PLoS One. 2013;8(6):e62973. 176. Kodama S, Tanaka S, Heianza Y, Fujihara K, Horikawa C, Shimano H, Saito K, Yamada N, Ohashi Y, and Sone H. Association between physical activity and risk of all-cause mortality and cardiovascular disease in patients with diabetes: a meta- analysis. Diabetes Care. 2013;36(2):471-9. 177. Sadarangani KP, Hamer M, Mindell JS, Coombs NA, and Stamatakis E. Physical activity and risk of all-cause and cardiovascular disease mortality in diabetic adults from Great Britain: pooled analysis of 10 population-based cohorts. Diabetes Care. 2014;37(4):1016-23. 178. Tanasescu M, Leitzmann MF, Rimm EB, Willett WC, Stampfer MJ, and Hu FB. Exercise type and intensity in relation to coronary heart disease in men. JAMA. 2002;288(16):1994-2000. 179. Woolf-May K, Kearney EM, Owen A, Jones DW, Davison RC, and Bird SR. The efficacy of accumulated short bouts versus single daily bouts of brisk walking in improving aerobic fitness and blood lipid profiles. Health Educ Res. 1999;14(6):803-15. 180. Woolf-May K, Kearney EM, Jones DW, Davison RC, Coleman D, and Bird SR. The effect of two different 18-week walking programmes on aerobic fitness, selected blood lipids and factor XIIa. J Sports Sci. 1998;16(8):701-10. 181. Murphy MH, and Hardman AE. Training effects of short and long bouts of brisk walking in sedentary women. Med Sci Sports Exerc. 1998;30(1):152-7. 182. Angadi SS, Weltman A, Watson-Winfield D, Weltman J, Frick K, Patrie J, and Gaesser GA. Effect of fractionized vs continuous, single-session exercise on blood pressure in adults. J Hum Hypertens. 2010;24(4):300-2. 183. Park S, Rink LD, and Wallace JP. Accumulation of physical activity leads to a greater blood pressure reduction than a single continuous session, in prehypertension. J Hypertens. 2006;24(9):1761-70. 184. Bouchard C, and Rankinen T. Individual differences in response to regular physical activity. Med Sci Sports Exerc. 2001;33(6 Suppl):S446-51; discussion S52-3. 185. Herrmann SD, Willis EA, Honas JJ, Lee J, Washburn RA, and Donnelly JE. Energy intake, nonexercise physical activity, and weight loss in responders and nonresponders: The Midwest Exercise Trial 2. Obesity (Silver Spring). 2015;23(8):1539-49. 186. Hautala AJ, Kiviniemi AM, Makikallio TH, Kinnunen H, Nissila S, Huikuri HV, and Tulppo MP. Individual differences in the responses to endurance and resistance training. Eur J Appl Physiol. 2006;96(5):535-42. 187. Sherrill DL, Kotchou K, and Quan SF. Association of physical activity and human sleep disorders. Arch Intern Med. 1998;158(17):1894-8. 188. Kim K, Uchiyama M, Okawa M, Liu X, and Ogihara R. An epidemiological study of insomnia among the Japanese general population. Sleep. 2000;23(1):41-7. 189. Youngstedt SD, Perlis ML, O'Brien PM, Palmer CR, Smith MT, Orff HJ, and Kripke DF. No association of sleep with total daily physical activity in normal sleepers. Physiol Behav. 2003;78(3):395-401.

190. Youngstedt SD. Ceiling and floor effects in sleep research. Sleep Med Rev. 2003;7(4):351-65. 191. Youngstedt SD, O'Connor PJ, and Dishman RK. The effects of acute exercise on sleep: a quantitative synthesis. Sleep. 1997;20(3):203-14. 192. Myllymaki T, Kyrolainen H, Savolainen K, Hokka L, Jakonen R, Juuti T, Martinmaki K, Kaartinen J, Kinnunen ML, and Rusko H. Effects of vigorous late- night exercise on sleep quality and cardiac autonomic activity. J Sleep Res. 2011;20(1 Pt 2):146-53. 193. Morita Y, Sasai-Sakuma T, and Inoue Y. Effects of acute morning and evening exercise on subjective and objective sleep quality in older individuals with insomnia. Sleep Med. 2017;34(200-8. 194. Oda S, and Shirakawa K. Sleep onset is disrupted following pre-sleep exercise that causes large physiological excitement at bedtime. Eur J Appl Physiol. 2014;114(9):1789-99. 195. Pescatello LS ed. ACSM's Guidelines for Exercise Testing and Prescription. Philadelphia, PA: Lippincott Williams & Wilkins; 2014. 196. Gulati M, Shaw LJ, Thisted RA, Black HR, Bairey Merz CN, and Arnsdorf MF. Heart rate response to exercise stress testing in asymptomatic women: the st. James women take heart project. Circulation. 2010;122(2):130-7. 197. Heyward VH, and Gibson A. Advanced Fitness Assessment and Exercise Prescription 7th Edition. Human Kinetics; 2014. 198. Borg G. Perceived exertion as an indicator of somatic stress. Scand J Rehabil Med. 1970;2(2):92-8. 199. Beltz NM, Gibson AL, Janot JM, Kravitz L, Mermier CM, and Dalleck LC. Graded Exercise Testing Protocols for the Determination of VO2max: Historical Perspectives, Progress, and Future Considerations. J Sports Med (Hindawi Publ Corp). 2016;2016(3968393. 200. Howley SKPET ed. Exercise Physiology: Theory and application to fitness and performance. New York, NY: McGraw-Hill; 2009. 201. Gleser MA. Effects of hypoxia and physical training on hemodynamic adjustments to one-legged exercise. J Appl Physiol. 1973;34(5):655-9. 202. Saltin B, Nazar K, Costill DL, Stein E, Jansson E, Essen B, and Gollnick D. The nature of the training response; peripheral and central adaptations of one-legged exercise. Acta Physiol Scand. 1976;96(3):289-305. 203. Pietrobelli A, Wang Z, and Heymsfield SB. Techniques used in measuring human body composition. Curr Opin Clin Nutr Metab Care. 1998;1(5):439-48. 204. Bays HE, Chapman RH, Grandy S, and Group SI. The relationship of body mass index to diabetes mellitus, hypertension and dyslipidaemia: comparison of data from two national surveys. Int J Clin Pract. 2007;61(5):737-47. 205. Lee SY, and Gallagher D. Assessment methods in human body composition. Curr Opin Clin Nutr Metab Care. 2008;11(5):566-72. 206. Lukaski HC. Soft tissue composition and bone mineral status: evaluation by dual- energy X-ray absorptiometry. J Nutr. 1993;123(2 Suppl):438-43. 207. Pietrobelli A, Formica C, Wang Z, and Heymsfield SB. Dual-energy X-ray absorptiometry body composition model: review of physical concepts. Am J Physiol. 1996;271(6 Pt 1):E941-51.

208. Hartig DE, and Henderson JM. Increasing hamstring flexibility decreases lower extremity overuse injuries in military basic trainees. Am J Sports Med. 1999;27(2):173-6. 209. Quante M, Kaplan ER, Rueschman M, Cailler M, Buxton OM, and Redline S. Practical considerations in using accelerometers to assess physical activity, sedentary behavior, and sleep. Sleep Health. 2015;1(4):275-84. 210. A Slater J, Botsis T, Walsh J, King S, Straker L, and Eastwood P. Assessing sleep using hip and wrist actigraphy: Hip and wrist actigraphy. Sleep and Biological Rhythms. 2015;13( 211. Mrosovsky N, Salmon PA, Menaker M, and Ralph MR. Nonphotic phase shifting in hamster clock mutants. J Biol Rhythms. 1992;7(1):41-9. 212. Mistlberger RE. Effects of daily schedules of forced activity on free-running rhythms in the rat. J Biol Rhythms. 1991;6(1):71-80. 213. Eastman CI, Hoese EK, Youngstedt SD, and Liu L. Phase-shifting human circadian rhythms with exercise during the night shift. Physiol Behav. 1995;58(6):1287-91. 214. Buxton OM, Lee CW, L'Hermite-Baleriaux M, Turek FW, and Van Cauter E. Exercise elicits phase shifts and acute alterations of melatonin that vary with circadian phase. Am J Physiol Regul Integr Comp Physiol. 2003;284(3):R714-24. 215. Edwards B, Waterhouse J, Atkinson G, and Reilly T. Exercise does not necessarily influence the phase of the circadian rhythm in temperature in healthy humans. J Sports Sci. 2002;20(9):725-32. 216. J. Wams E, Riede S, van der Laan I, ten Bulte T, and Hut R. 2017:395-404. 217. Van Reeth O, Sturis J, Byrne MM, Blackman JD, L'Hermite-Baleriaux M, Leproult R, Oliner C, Refetoff S, Turek FW, and Van Cauter E. Nocturnal exercise phase delays circadian rhythms of melatonin and thyrotropin secretion in normal men. Am J Physiol. 1994;266(6 Pt 1):E964-74. 218. Baehr EK, Eastman CI, Revelle W, Olson SH, Wolfe LF, and Zee PC. Circadian phase-shifting effects of nocturnal exercise in older compared with young adults. Am J Physiol Regul Integr Comp Physiol. 2003;284(6):R1542-50. 219. Buxton OM, Frank SA, L'Hermite-Baleriaux M, Leproult R, Turek FW, and Van Cauter E. Roles of intensity and duration of nocturnal exercise in causing phase delays of human circadian rhythms. Am J Physiol. 1997;273(3 Pt 1):E536-42. 220. Youngstedt SD, Kline CE, Elliott JA, Zielinski MR, Devlin TM, and Moore TA. Circadian Phase-Shifting Effects of Bright Light, Exercise, and Bright Light + Exercise. J Circadian Rhythms. 2016;14(2. 221. Youngstedt SD, Elliott JA, and Kripke DF. Human circadian phase-response curves for exercise. J Physiol. 2019. 222. Buxton OM, L'Hermite-Baleriaux M, Hirschfeld U, and Cauter E. Acute and delayed effects of exercise on human melatonin secretion. J Biol Rhythms. 1997;12(6):568-74. 223. Skrinar GS, Bullen BA, Reppert SM, Peachey SE, Turnbull BA, and McArthur JW. Melatonin response to exercise training in women. J Pineal Res. 1989;7(2):185-94.

224. Tranel HR, Schroder EA, England J, Black WS, Bush H, Hughes ME, Esser KA, and Clasey JL. Physical activity, and not fat mass is a primary predictor of circadian parameters in young men. Chronobiol Int. 2015;32(6):832-41. 225. Satlin A, Volicer L, Stopa EG, and Harper D. Circadian locomotor activity and core-body temperature rhythms in Alzheimer's disease. Neurobiol Aging. 1995;16(5):765-71. 226. Breen DP, Vuono R, Nawarathna U, Fisher K, Shneerson JM, Reddy AB, and Barker RA. Sleep and circadian rhythm regulation in early Parkinson disease. JAMA Neurol. 2014;71(5):589-95. 227. Hart PD, Benavidez G, and Erickson J. Meeting Recommended Levels of Physical Activity in Relation to Preventive Health Behavior and Health Status Among Adults. J Prev Med Public Health. 2017;50(1):10-7. 228. Marchant EG, Watson NV, and Mistlberger RE. Both neuropeptide Y and serotonin are necessary for entrainment of circadian rhythms in mice by daily treadmill running schedules. J Neurosci. 1997;17(20):7974-87. 229. Wolff G, and Esser KA. Scheduled exercise phase shifts the circadian clock in skeletal muscle. Med Sci Sports Exerc. 2012;44(9):1663-70. 230. van Moorsel D, Hansen J, Havekes B, Scheer F, Jorgensen JA, Hoeks J, Schrauwen-Hinderling VB, Duez H, Lefebvre P, Schaper NC, et al. Demonstration of a day-night rhythm in human skeletal muscle oxidative capacity. Mol Metab. 2016;5(8):635-45. 231. Zambon AC, McDearmon EL, Salomonis N, Vranizan KM, Johansen KL, Adey D, Takahashi JS, Schambelan M, and Conklin BR. Time- and exercise-dependent gene regulation in human skeletal muscle. Genome Biol. 2003;4(10):R61. 232. Fernandes AL, Lopes-Silva JP, Bertuzzi R, Casarini DE, Arita DY, Bishop DJ, and Lima-Silva AE. Effect of time of day on performance, hormonal and metabolic response during a 1000-M cycling time trial. PLoS One. 2014;9(10):e109954. 233. Bernard T, Giacomoni M, Gavarry O, Seymat M, and Falgairette G. Time-of-day effects in maximal anaerobic leg exercise. Eur J Appl Physiol Occup Physiol. 1998;77(1-2):133-8. 234. Gray SR, De Vito G, Nimmo MA, Farina D, and Ferguson RA. Skeletal muscle ATP turnover and muscle fiber conduction velocity are elevated at higher muscle temperatures during maximal power output development in humans. Am J Physiol Regul Integr Comp Physiol. 2006;290(2):R376-82. 235. Gauthier A, Davenne D, Martin A, Cometti G, and Van Hoecke J. Diurnal rhythm of the muscular performance of elbow flexors during isometric contractions. Chronobiol Int. 1996;13(2):135-46. 236. Kuusmaa-Schildt M, Liukkonen J, Vuong MK, Nyman K, Hakkinen K, and Hakkinen A. Effects of morning vs. evening combined strength and endurance training on physical performance, sleep and well-being. Chronobiol Int. 2019;36(6):811-25. 237. Ezagouri S, Zwighaft Z, Sobel J, Baillieul S, Doutreleau S, Ladeuix B, Golik M, Verges S, and Asher G. Physiological and Molecular Dissection of Daily Variance in Exercise Capacity. Cell Metab. 2019.

238. Sato S, Basse AL, Schonke M, Chen S, Samad M, Altintas A, Laker RC, Dalbram E, Barres R, Baldi P, et al. Time of Exercise Specifies the Impact on Muscle Metabolic Pathways and Systemic Energy Homeostasis. Cell Metab. 2019. 239. DeCoursey PJ, and Krulas JR. Behavior of SCN-lesioned chipmunks in natural habitat: a pilot study. J Biol Rhythms. 1998;13(3):229-44. 240. Hurd MW, and Ralph MR. The significance of circadian organization for longevity in the golden hamster. J Biol Rhythms. 1998;13(5):430-6. 241. Woelfle MA, Ouyang Y, Phanvijhitsiri K, and Johnson CH. The adaptive value of circadian clocks: an experimental assessment in cyanobacteria. Curr Biol. 2004;14(16):1481-6. 242. Vera B, Dashti HS, Gomez-Abellan P, Hernandez-Martinez AM, Esteban A, Scheer F, Saxena R, and Garaulet M. Modifiable lifestyle behaviors, but not a genetic risk score, associate with metabolic syndrome in evening chronotypes. Sci Rep. 2018;8(1):945. 243. Yu JH, Yun CH, Ahn JH, Suh S, Cho HJ, Lee SK, Yoo HJ, Seo JA, Kim SG, Choi KM, et al. Evening chronotype is associated with metabolic disorders and body composition in middle-aged adults. J Clin Endocrinol Metab. 2015;100(4):1494-502. 244. Baron KG, Reid KJ, Kim T, Van Horn L, Attarian H, Wolfe L, Siddique J, Santostasi G, and Zee PC. Circadian timing and alignment in healthy adults: associations with BMI, body fat, caloric intake and physical activity. Int J Obes (Lond). 2017;41(2):203-9. 245. Duffy JF, and Wright KP, Jr. Entrainment of the human circadian system by light. J Biol Rhythms. 2005;20(4):326-38. 246. Duffy JF, Kronauer RE, and Czeisler CA. Phase-shifting human circadian rhythms: influence of sleep timing, social contact and light exposure. J Physiol. 1996;495 ( Pt 1)(289-97. 247. Moore RY, and Lenn NJ. A retinohypothalamic projection in the rat. J Comp Neurol. 1972;146(1):1-14. 248. La Fleur SE. Daily rhythms in glucose metabolism: suprachiasmatic nucleus output to peripheral tissue. J Neuroendocrinol. 2003;15(3):315-22. 249. Kalsbeek A, Fliers E, Romijn JA, La Fleur SE, Wortel J, Bakker O, Endert E, and Buijs RM. The suprachiasmatic nucleus generates the diurnal changes in plasma leptin levels. Endocrinology. 2001;142(6):2677-85. 250. La Fleur SE, Kalsbeek A, Wortel J, and Buijs RM. A suprachiasmatic nucleus generated rhythm in basal glucose concentrations. J Neuroendocrinol. 1999;11(8):643-52. 251. Hut RA, Mrosovsky N, and Daan S. Nonphotic entrainment in a diurnal mammal, the European ground squirrel (Spermophilus citellus). J Biol Rhythms. 1999;14(5):409-19. 252. Woods RA. Sports and Exercise. US BUREAU OF LABOR STATISTICS. 2017. 253. Perreau-Lenz S, Kalsbeek A, Garidou ML, Wortel J, van der Vliet J, van Heijningen C, Simonneaux V, Pevet P, and Buijs RM. Suprachiasmatic control of melatonin synthesis in rats: inhibitory and stimulatory mechanisms. Eur J Neurosci. 2003;17(2):221-8.

254. Lewy AJ, and Sack RL. The dim light melatonin onset as a marker for circadian phase position. Chronobiol Int. 1989;6(1):93-102. 255. Wright KP, Jr., Gronfier C, Duffy JF, and Czeisler CA. Intrinsic period and light intensity determine the phase relationship between melatonin and sleep in humans. J Biol Rhythms. 2005;20(2):168-77. 256. van der Meijden WP, Van Someren JL, Te Lindert BH, Bruijel J, van Oosterhout F, Coppens JE, Kalsbeek A, Cajochen C, Bourgin P, and Van Someren EJ. Individual Differences in Sleep Timing Relate to Melanopsin-Based Phototransduction in Healthy Adolescents and Young Adults. Sleep. 2016;39(6):1305-10. 257. McMahon DM, Burch JB, Youngstedt SD, Wirth MD, Hardin JW, Hurley TG, Blair SN, Hand GA, Shook RP, Drenowatz C, et al. Relationships between chronotype, social jetlag, sleep, obesity and blood pressure in healthy young adults. Chronobiol Int. 2019;36(4):493-509. 258. Duffy JF, and Czeisler CA. Effect of Light on Human Circadian Physiology. Sleep Med Clin. 2009;4(2):165-77. 259. Vetter C, Devore EE, Wegrzyn LR, Massa J, Speizer FE, Kawachi I, Rosner B, Stampfer MJ, and Schernhammer ES. Association Between Rotating Night Shift Work and Risk of Coronary Heart Disease Among Women. JAMA. 2016;315(16):1726-34. 260. Jankowiak S, Backe E, Liebers F, Schulz A, Hegewald J, Garthus-Niegel S, Nubling M, Blankenberg S, Pfeiffer N, Lackner KJ, et al. Current and cumulative night shift work and subclinical atherosclerosis: results of the Gutenberg Health Study. Int Arch Occup Environ Health. 2016;89(8):1169-82. 261. Vetter C, Devore EE, Ramin CA, Speizer FE, Willett WC, and Schernhammer ES. Mismatch of Sleep and Work Timing and Risk of Type 2 Diabetes. Diabetes Care. 2015;38(9):1707-13. 262. Youngstedt SD, Kripke DF, and Elliott JA. Circadian phase-delaying effects of bright light alone and combined with exercise in humans. Am J Physiol Regul Integr Comp Physiol. 2002;282(1):R259-66. 263. Sletten TL, Vincenzi S, Redman JR, Lockley SW, and Rajaratnam SM. Timing of sleep and its relationship with the endogenous melatonin rhythm. Front Neurol. 2010;1(137. 264. Riebe D, Ehrman JK, Liguori G, and Magal M. ACSM's guidelines for exercise testing and prescription. 2018. 265. Lohman TG, Roche AF, and Martorell R. Anthropometric standardization reference manual. Champaign, IL: Human Kinetics Books; 1988. 266. Loewen A, Siemens A, and Hanly P. Sleep disruption in patients with sleep apnea and end-stage renal disease. J Clin Sleep Med. 2009;5(4):324-9. 267. Wright KP, Jr., Badia P, Myers BL, Plenzler SC, and Hakel M. Caffeine and light effects on nighttime melatonin and temperature levels in sleep-deprived humans. Brain Res. 1997;747(1):78-84. 268. Rupp TL, Acebo C, and Carskadon MA. Evening alcohol suppresses salivary melatonin in young adults. Chronobiol Int. 2007;24(3):463-70.

100

269. Sae-Teaw M, Johns J, Johns NP, and Subongkot S. Serum melatonin levels and antioxidant capacities after consumption of pineapple, orange, or banana by healthy male volunteers. J Pineal Res. 2013;55(1):58-64. 270. Lohman T RA, Matorell R. Anthropometric standardization reference manual. Champaign, IL: Human Kinetics; 1988. 271. American College of Sports Medicine. ACSM's guideliens for exercise testing and prescription. Philadelphia: Lippincott Williams & Wilkins; 2010.

101

Vita J. Matthew Thomas Education INSTITUTION AND DEGREE COMPLETION FIELD OF STUDY LOCATION DATE University of Kentucky, B.S 05/2011 Biology Lexington, KY University of Kentucky, M.S. 05/2015 Exercise Physiology Lexington, KY

Professional Positions 2019- Clinical Research Coordinator, Department of Internal Medicine- Gastroenterology, University of Kentucky, Lexington, KY 2019 Graduate Teaching Assistant, Microbiology Lab, Dept. of Biology, University of Kentucky, Lexington, KY 2016-2018 TL1 Predoctoral Fellow, University of Kentucky, Lexington, KY 2016 Graduate Teaching Assistant, Exercise Physiology Lab, Dept. of Kinesiology and Health Promotion, University of Kentucky, Lexington, KY 2013-2016 Graduate Teaching Assistant, Microbiology Lab, Dept. of Biology, University of Kentucky, Lexington, KY 2012 Graduate Teaching Assistant, General Biology Lab, Dept. of Biology, University of Kentucky, Lexington, KY

Scholastic and Professional Honors 2019 Doctoral Student Research Award, Southeast American College of Sports Medicine Annual Meeting, Greenville, SC 2019 Department of Kinesiology and Health Promotion Travel Award 2018 Arvle and Ellen Turner Thacker Dissertation Support Award 2018 Department of Kinesiology and Health Promotion Travel Award 2018 Frank G. and Elizabeth D. Dickey Graduate Fellowship in Education 2017 Hackensmith Outstanding Graduate Student Award Nominee, Department of Kinesiology and Health Promotion 2017 Arvle and Ellen Turner Thacker Dissertation Support Award 2016-2018 TL1TR001997 – National Center for Advancing Translational Sciences National Institutes of Health 2015 Department of Kinesiology and Health Promotion Travel Award 2010 Summer Undergraduate Research and Creativity Grant

102

Publications J.J. Krupa, and J. Matthew Thomas. Is the common teasel (Dipsacus Fullonum) carnivorous or was Francis Darwin wrong? Botany. 97(6): 321-328, 2019. J. Matthew Thomas, M. Pohl, R. Shapiro, J. Keeler, and M.G. Abel. Effect of load carriage on tactical performance in special weapons and tactics operators. Journal of Strength and Conditioning. 32(2): 554-564, 2018.

Abstracts J. M. Thomas, J. S. Pendergast, W. S. Black, P. A. Kern, J. L. Clasey. Circadian Phase is Associated with Self-reported Chronotype in Young, Sedentary Adults. American College of Sports Medicine 66th annual meeting, Orlando, Florida, May 30, 2019.

J. M. Thomas, J. S. Pendergast, W. S. Black, P. A. Kern, J. L. Clasey. Fat Mass, and not Heart Rate Recovery is Associated with Cardiorespiratory Fitness in Young, Sedentary Adults. University of Kentucky Center for Clinical and Translational Science 14th Annual Spring Conference. Lexington, Kentucky. April 15, 2019.

J. M. Thomas, J. S. Pendergast, W. S. Black, P. A. Kern, J. L. Clasey. Fat Mass, and not Heart Rate Recovery is Associated with Cardiorespiratory Fitness in Young, Sedentary Adults. College of Education and Human Development Spring Research Conference. Lexington, Kentucky. March 2, 2019.

J. M. Thomas, J. S. Pendergast, W. S. Black, P. A. Kern, J. L. Clasey. Circadian Phase is Associated with Self-reported Chronotype in Young, Sedentary Adults. Southeast American College of Sports Medicine annual meeting, Greenville, South Carolina, February 15, 2019.

J. M. Thomas, J. S. Pendergast, W. S. Black, P. A. Kern, J. L. Clasey. Fat Mass, and not Heart Rate Recovery is Associated with Cardiorespiratory Fitness in Young, Sedentary Adults. 21st Annual Saha Cardiovascular Research Day. Lexington, Kentucky, September 21, 2018.

J. M. Thomas, J. S. Pendergast, W. S. Black, P. A. Kern, J. L. Clasey. Fat Mass, and not Heart Rate Recovery is Associated with Cardiorespiratory Fitness in Young, Sedentary Adults. American College of Sports Medicine 65th Annual Meeting, Minneapolis, Minnesota. May 31, 2018.

J. M. Thomas, J. L. Clasey, W. S. Black, P. A. Kern, J. S. Pendergast. Effects of Timed Exercise on Circadian Rhythms in Humans. Society for Research on Biological Rhythms Annual Conference. Amelia Island, FL. May 13, 2018.

J. M. Thomas, J. S. Pendergast, W. S. Black, P. A. Kern, J. L. Clasey. Metabolic Benefits of Timed Exercise. University of Kentucky Center for Clinical and Translational Science 13th Annual Spring Conference. Lexington, Kentucky, April 20, 2018.

103

J. M. Thomas, J. S. Pendergast, W. S. Black, P. A. Kern, J. L. Clasey. Reducing Circadian Misalignment with Timed Exercise. College of Education and Human Development Spring Research Conference. Louisville, Kentucky. March 24, 2018.

J. M. Thomas, J. S. Pendergast, W. S. Black, P. A. Kern, J. L. Clasey. Metabolic Benefits of Timed Exercise. University of Kentucky 7th Annual Obesity and Diabetes Research Day, Lexington, Kentucky. May 18, 2017.

J. M. Thomas, J. S. Pendergast, W. S. Black, P. A. Kern, J. L. Clasey. Metabolic Benefits of Timed Exercise. Association for Clinical and Translational Science Annual Meeting Washington D.C., April 20, 2017.

J. M. Thomas, J. S. Pendergast, W. S. Black, P. A. Kern, J. L. Clasey. Metabolic Benefits of Timed Exercise. University of Kentucky Center for Clinical and Translational Science 12th Annual Spring Conference. Lexington, Kentucky, March 30, 2017.

J. M. Thomas, M. Pohl, R. Shapiro, J. Keeler, and M.G. Abel. Effect of load carriage on tactical physical ability and marksmanship in SWAT operators. National Strength and Conditioning Association National Conference. Orlando, Florida. July 9, 2015.

104