Copyright is owned by the Author of the thesis.  Permission is given for 
a copy to be downloaded by an individual for the purpose of research and 
private study only.  The thesis may not be reproduced elsewhere without 
the permission of the Author. 
 



 

 

The effect of creatine monohydrate loading on 

cognitive performance, mood, sleepiness, and 

perceived workload following sleep restriction 

 

A thesis presented in partial fulfilment of the requirements for the degree of 

Master of Science in Nutrition and Dietetics 

 

 

Massey University, Albany 

 New Zealand 

 

 

 

 Juliette Victoire Constance Janvresse 

 2023 



i 
 

Abstract 

Background: Sleep restriction (SR; i.e. not getting enough sleep per 24 hours for one or 

multiple nights) impairs cognitive performance and mood, which can compromise capability 

and safety. SR-related performance deficits may be underpinned by disruptions to brain 

energy metabolism. Sleep loss appears to reduce brain phosphorylcreatine (PCr; also called 

creatine phosphate and phosphocreatine), which supports the regeneration of adenosine 

triphosphate (ATP) from adenosine diphosphate (ADP). Since creatine supplementation 

increases tissue PCr, including in the brain, it may mitigate possible SR-related cognitive 

impairments.  

Aim: To assess the effect of a 7-day creatine monohydrate (CrM) loading protocol on cognitive 

performance, mood, subjective sleepiness, and perceived workload following one night of SR 

(3 hours time-in-bed; TIB). 

Methods: A double-blinded, randomised, placebo-controlled cross-over trial was conducted 

with 7 healthy participants (3 men and 4 women); age, 28.0 ± 4.6 years (range, 22-35 years). 

Participants ingested 20 gday-1 of CrM or placebo (i.e. tapioca) divided into 4 equal 5 g doses 

for 7 days followed by a single SR night (3 hours TIB). Cognitive performance (10-minute 

Psychomotor Vigilance Task; PVT), mood (fatigue and vigour), subjective sleepiness, and 

perceived workload were measured following ~8 hours of sleep prior to creatine 

supplementation and following the creatine loading protocol and SR. A minimum of a 5-week 

wash-out period separated the two trial arms. Linear mixed effects models were employed to 

analyse data.   

Results: SR slowed mean reciprocal reaction time (p = 0.01) and mean 10% fastest reaction 

time (p = 0.02), increased fatigue (p = 0.004), subjective sleepiness (p = 0.009), workload (p = 

0.02), and reduced vigour (p <0.001); however, it did not alter the number of lapses (p = 0.60) 

and mean 10% slowest reaction time (p = 0.60). CrM loading had no effect on cognitive 

performance, mood, subjective sleepiness, or workload. 



ii 
 

Conclusion: A single night of SR (3 hours TIB) negatively affected aspects of PVT performance, 

mood, subjective sleepiness, and perceived workload; however, CrM loading (20 gday-1 for 7 

days) did not appear to mitigate the effects of SR.  

Key words: sleep loss; psychomotor vigilance task; randomised cross-over trial 

  



iii 
 

Acknowledgement 

This thesis has been a very challenging journey, and I want to express my gratitude toward all 

who were directly and indirectly involved in this study. 

First, I would like to thank my supervisors Dr David Shaw and Dr Margo van den Berg. Although 

the research has not always gone to plan, your expertise, invaluable feedback and support 

throughout this process will always be appreciated. Thank you for the time and the effort you 

put into this research. 

I would like to thank all the participants who generously offered their time and accepted to 

take part of this challenging protocol. Thank you for making this study possible.  

To my mum, dad, sister, Manon and my best friend, Mathilde, who despite living on the other 

side of the planet, reassured me and gave me strength when I needed it. Thank you for your 

continued support and encouragement.  

Last but not least, I want to thank my wonderful partner, Sumit. Without you, I doubt I would 

have managed to finish this thesis, and I can’t thank you enough for all the financial, 

emotional, and technical support you provided me over the intense months of work. Thank 

you for always believing in me. 

 

  



iv 
 

Table of Contents 

 

Abstract ....................................................................................................................................... i 

Acknowledgement..................................................................................................................... iii 

Table of Contents ...................................................................................................................... iv 

List of Figures .............................................................................................................................vii 

List of Tables ............................................................................................................................. viii 

1 Introduction ........................................................................................................................ 1 

1.1 Background .................................................................................................................. 1 

1.2 Research aims .............................................................................................................. 3 

1.3 Objectives .................................................................................................................... 3 

1.4 Hypothesis ................................................................................................................... 3 

1.5 Structure of the thesis ................................................................................................. 3 

1.6 Contributors to the research ....................................................................................... 4 

2 Literature Review ................................................................................................................ 5 

2.1 Sleep ............................................................................................................................ 5 

2.1.1 Structure of sleep ................................................................................................. 5 

2.1.2 Measuring sleep ................................................................................................... 6 

2.1.3 Sleep regulation.................................................................................................... 6 

2.1.4 Sleep restriction ................................................................................................... 8 

2.1.5 Summary ............................................................................................................ 11 

2.2 What is creatine? ....................................................................................................... 11 

2.2.1 Sources of creatine ............................................................................................. 13 

2.2.2 Optimal protocol for creatine supplementation ................................................ 14 

2.2.3 Safety of creatine supplementation ................................................................... 14 

2.2.4 Creatine supplementation and cognition .......................................................... 15 

2.2.5 Creatine supplementation, sleep loss and cognition ......................................... 15 

2.3 Conclusion.................................................................................................................. 18 

2.3.1 Comment from the author ................................................................................. 18 

3 Research Manuscript ........................................................................................................ 19 

3.1 Abstract ...................................................................................................................... 20 

3.2 Introduction ............................................................................................................... 21 



v 
 

3.3 Material and Methods ............................................................................................... 22 

3.3.1 Study overview ................................................................................................... 22 

3.3.2 Eligibility criteria ................................................................................................. 22 

3.3.3 Experimental procedures ................................................................................... 23 

3.3.4 Dietary supplementation ................................................................................... 24 

3.3.5 Testing protocol.................................................................................................. 24 

3.3.6 Dietary monitoring ............................................................................................. 26 

3.3.7 Sleep monitoring ................................................................................................ 26 

3.3.8 Statistical Analysis .............................................................................................. 26 

3.3.9 Comment from the author ................................................................................. 27 

3.4 Results ........................................................................................................................ 27 

3.4.1 Participants ......................................................................................................... 27 

3.4.2 Supplement and dietary compliance ................................................................. 28 

3.4.3 Sleep duration .................................................................................................... 29 

3.4.4 PVT performance ................................................................................................ 30 

3.4.5 Mood, subjective sleepiness, and workload ...................................................... 31 

3.5 Discussion .................................................................................................................. 33 

3.7.1 Effect of sleep restriction ......................................................................................... 33 

3.5.1 Effect of creatine supplementation following sleep restriction ........................ 34 

3.5.2 Brain creatine stores .......................................................................................... 34 

3.5.3 Study limitations ................................................................................................. 35 

3.5.4 Conclusion .......................................................................................................... 35 

4 Conclusion and recommendations ................................................................................... 37 

4.1 Study summary .......................................................................................................... 37 

4.2 Main findings ............................................................................................................. 38 

4.2.1 Hypothesis 1: A single sleep restriction night (3 hours TIB) will negatively affect 

cognitive performance, mood, subjective sleepiness, and perceived workload. ............ 38 

4.2.2 Hypothesis 2: CrM loading will mitigate the adverse effect of a single night of 

sleep restriction night (3 hours TIB) on cognitive performance, mood, subjective 

sleepiness, and perceived workload. ................................................................................ 38 

4.3 Study strengths .......................................................................................................... 38 

4.4 Study limitations ........................................................................................................ 40 

4.5 Final recommendations ............................................................................................. 41 



vi 
 

4.5.1 Impact ................................................................................................................. 41 

4.5.2 Recommendations ............................................................................................. 41 

5 References ........................................................................................................................ 43 

Appendices ............................................................................................................................... 50 

Appendix A: Participant information sheet .......................................................................... 50 

Appendix B: Participant calendar ......................................................................................... 60 

Appendix C: Caffeine handout .............................................................................................. 61 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 



vii 
 

List of Figures 

Figure 2.1: The Two-Process Model of Sleep Regulation ........................................................... 8 

Figure 2.2: Creatine molecule .................................................................................................. 12 

Figure 2.3: Lohman reaction .................................................................................................... 12 

Figure 2.4: Creatine synthesis. ................................................................................................. 13 

Figure 3.1: An overview of the experimental design. .............................................................. 24 

Figure 3.2: Participant selection ............................................................................................... 28 

Figure 3.3: Psychomotor vigilance task performance at baseline and after SR following 7 days 

of CrM supplementation. ......................................................................................................... 31 

Figure 3.4: Mood, subjective sleepiness, and workload at baseline and after SR following 7 

days of CrM supplementation .................................................................................................. 32 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 



viii 
 

List of Tables 

Table 1.1: Researchers' Contributions. ...................................................................................... 4 

Table 2.1: The effect of creatine supplementation on cognition following sleep loss. ........... 17 

Table 3.1: Sleep duration for baseline, 7-day intervention and SR for CrM and placebo ....... 30 

Table 3.1: Sleep duration for baseline, 7-day intervention and SR for CrM and placebo. ...... 30 

Table 3.2: Duration of wakefulness prior to each testing time. .............................................. 30 

Table 3.3: P-values of main effects and interactions for PVT variables ................................... 31 

Table 3.4: P-values of main effects and interactions for mood, subjective sleepiness, and 

workload……………………………………………………………………………………………………………………………..33 

 



1 
 

1 Introduction 

1.1 Background 

Obtaining an adequate amount of sleep (7 to 9 hours for most adults) is essential for optimal 

well-being and performance (Hirshkowitz et al., 2015; Watson et al., 2015). However, due to 

work schedules, long duty hours, personal or family commitments, social and sporting 

engagements, as well as individuals’ life choices or medical conditions (Banks et al., 2017; 

McLellan et al., 2016; Rupp et al., 2012), a significant proportion (37%) of the population 

regularly fail to achieve sleep recommendations (Lee & Sibley, 2019). Sleep restriction (SR; i.e. 

not getting enough sleep per 24 hours for one or multiple nights) increases fatigue, sleepiness 

and impairs cognitive performance and mood (Banks et al., 2017; Lowe et al., 2017). 

Unintentional human errors are estimated to contribute between 30% to 90% of all serious 

incidents across industries (Reason, 1990), with fatigue and sleepiness being primary causes 

(Dinges, 1995). This is a concern in work areas affected by shift work, such as military, 

healthcare and transport, as it can compromise capability and safety (Caldwell et al., 2019; 

McLellan et al., 2016). Strategies to mitigate the detrimental effects of inadequate sleep on 

cognition are required. 

The brain is a highly metabolic organ, consuming ~20-25% of the resting metabolic rate in 

humans (Bailey, 2019), and therefore requires a constant energy supply. Following sleep loss, 

brain energy metabolism is disrupted, which likely underpins some cognitive impairments 

(Engle-Friedman, 2014; Thomas et al., 2000). Glucose is the primary energy source for the 

brain (Dienel, 2019), but it can also utilise alternative non-oxidisable substrates, such as 

phosphorylcreatine (PCr) (Greenhaff, 2001; Walker, 1979), to generate adenosine 

triphosphate (ATP), the energy currency of the cell. ATP is hydrolysised to adenosine 

diphosphate (ADP) and inorganic phosphate to produce energy (Greenhaff, 2001). The re-

synthesis of ATP is modulated by PCr due to its ability to donate a phosphate group to 

adenosine di-phosphate (ADP) via the enzyme creatine kinase (Greenhaff, 2001; Walker, 

1979). PCr is synthesised by phosphorylating creatine using the same enzyme, creatine kinase, 

in a reverse reaction (Greenhaff, 2001; Walker, 1979). The availability of PCr, however, can 

decline when energy demand is increased, such as during exhausting mental tasks, or 

following sleep loss (Dworak et al., 2010; Lyoo et al., 2003; Plante et al., 2014). Increasing brain 



2 
 

creatine, and therefore PCr stores, is a possible strategy to mitigate cognitive impairments 

following sleep loss, which can be achieved by supplementing oral intake of creatine (Dworak 

et al., 2017; McMorris et al., 2007; McMorris et al., 2006).  

Creatine is a naturally occurring guanidine compound involved in storage and transport of 

energy (Walker, 1979). Creatine is mostly found in muscles due to their large mass, but it is 

also stored in the brain (Rae & Bröer, 2015). Creatine comes from dietary sources (mostly 

meat and fish) and endogenous synthesis (Andres et al., 2008; East, 2002; Wallimann et al., 

2007). Creatine can also be ingested as a supplement, usually in the form of creatine 

monohydrate (CrM), to augment tissue concentrations (Kreider et al., 2017). CrM 

supplementation is safe (European Food Safety Acency, 2004; Kim et al., 2011) and typically 

ingested over several days-to-weeks (i.e. CrM loading). Whereas dietary creatine does not 

seem to influence brain creatine stores (Kaviani et al., 2020; Solis et al., 2014), CrM loading 

protocols have demonstrated to increase brain creatine and PCr stores by 4.7 to 11.5% 

(depending on the brain areas), and 9.2%, respectively (Dechent et al., 1999; Turner et al., 

2015). The benefits of creatine supplementation on cognitive performance and mood seem 

most pronounced under physiological stress, such as demanding cognitive tasks (van Cutsem 

et al., 2020; Watanabe et al., 2002), sleep deprivation (SD; i.e. no sleep) (McMorris et al., 2007; 

McMorris et al., 2006) and hypoxia (Turner et al., 2015). Mitigating cognitive fatigue with 

creatine supplementation may also reduce perceived workload associated with demanding 

cognitive tasks, yet such effects are still to be elucidated (van Cutsem et al., 2020). 

Whether creatine supplementation can support cognition following SR remains uncertain, but 

as SR is more common within most communities than total SD, findings would be highly 

relevant. A single night of SR (i.e. ~5 hours time-in-bed; TIB) impairs performance across 

multiple cognitive domains (Lowe et al., 2017), particularly simple and monotonous tasks, 

such as those assessing vigilance (Durmer & Dinges, 2005). Vigilance can be measured using 

the Psychomotor Vigilance Task (PVT) (Dinges & Powell, 1985), which is a simple, valid, and 

reliable measure of the effects of sleep loss, and circadian rhythm in cognitive performance 

(Lim & Dinges, 2008, 2010). Supporting measures of subjective sleepiness, mood (fatigue and 

vigour), and perceived workload would complement observations on cognitive performance. 

To the author’s knowledge, however, the effects of creatine supplementation on cognition 

following SR have not yet been investigated. 



3 
 

1.2 Research aims 

The aim of this research thesis is to assess the effect of a 7-day CrM loading protocol on 

psychomotor vigilance performance, mood (vigour and fatigue), subjective sleepiness, and 

perceived workload following one night of SR (3 hours TIB) in healthy men and women. 

1.3 Objectives 

The objectives of this research thesis are to examine whether 7 days of ingesting 20 gday-1 of 

CrM affects aspects of cognition following a single night of SR (3 hours TIB), including: 

- Simple cognitive performance (vigilance and sustained attention), using a 10-minute 

version of the PVT (PVT-192). 

- Mood (fatigue and vigour), using a the validated computerised ANAM® mood scale. 

- Subjective sleepiness, using the Stanford Sleepiness Scale. 

- Perceived workload, using the NASA Task Load Index (NASA-TLX) 

1.4 Hypothesis 

1. A single sleep restriction night (3 hours TIB) will negatively affect cognitive 

performance, mood, subjective sleepiness, and perceived workload. 

2. CrM loading will mitigate the adverse effects of a single night of sleep restriction (3 

hours TIB) on cognitive performance, mood, subjective sleepiness, and perceived 

workload. 

1.5 Structure of the thesis 

This thesis consists of four chapters. The first chapter is an introduction identifying the 

purpose of the study. Chapter two is a narrative literature review on sleep, the effect of SR on 

cognition and brain energy metabolism, creatine supplementation, and the effects of creatine 

supplementation on cognition and possible implications for SR. Chapter three is the research 

manuscript, providing a complete presentation of the study, followed by the discussion of the 

findings and the conclusion. The fourth and final chapter is a conclusion, including the 

strengths and limitations of the study, potential application of observations and suggestions 

for future research. Appendices, including participant information sheet, participant calendar, 

and caffeine handout are added at the end of the thesis.  



4 
 

1.6 Contributors to the research 

Table 1.1: Researchers' Contributions. 

Researchers Contributions 

Dr David Shaw, PhD 

Academic main supervisor 

Main contributor to study design, ethics 

application, participant recruitment, study 

implementation and data collection; creation of the 

R script for data analyses; academic support of the 

thesis write up. 

Dr Margo van den Berg, PhD 

Academic co-supervisor 

Provision of actigraphs and PVT devices; scoring of 

participants’ actigraphy data; interpretation of 

results; academic support of the thesis write up. 

Juliette Janvresse, Student Dietitian Review of literature; creation of participant 

resource (i.e. calendar); study implementation and 

data collection; data processing and interpretation; 

author of the thesis. 

Associate Professor Nick Gant, PhD 

 

Guidance on the use of creatine supplementation 

and use of the Exercise Neurometabolism 

Laboratory, University of Auckland, to create the 

creatine and placebo supplements. 

Peter Bloomfield, PhD Candidate Data collection. 

 

 

 

 

  



5 
 

2 Literature Review 

2.1 Sleep 

Sleep is essential for optimal health and performance (Hirshkowitz et al., 2015; Watson et al., 

2015). Sleep is “a reversible behavioural state of perceptual disengagement from and 

unresponsiveness to the environment” (Carskadon & Dement, 2017, p. 1). Although the 

function of sleep is not completely understood, sleep seems to be required for brain 

processing, memory consolidation and restoring brain energy stores (Inoué et al., 1995; 

Stickgold, 2005; Xie et al., 2013). There are large individual differences in sleep need (Dennis 

et al., 2017; Van Dongen et al., 2004) with some individuals requiring up to 10 hours to 

function optimally, whereas others need as few as 6 hours (Hirshkowitz et al., 2015). However, 

the recommended optimal sleep duration for adults (18 to 64 years old) is between 7 to 9 

hours per night on a regular basis (Hirshkowitz et al., 2015; Watson et al., 2015). 

Unfortunately, lack of sleep is commonly experienced in our society due to social drivers and 

work demands (Chatzitheochari & Arber, 2009; Lee & Sibley, 2019). Indeed, in a recent survey, 

37% of adults in New Zealand reported achieving less than the minimum recommended sleep 

duration (Lee & Sibley, 2019). People working in occupations and industries such as the 

military, medical care, transport, search and rescue, are also at risk of inadequate sleep due 

to work schedules and long duty hours overlapping with the person’s normal sleep periods 

(McLellan et al., 2016). Inadequate sleep increases fatigue and sleepiness, which impairs 

cognitive performance and vigilance, thus increasing the risk of errors and accidents (Banks et 

al., 2017; Caldwell et al., 2019; van Dongen et al., 2017). Strategies mitigating the adverse 

effects of sleep loss are therefore warranted. 

2.1.1 Structure of sleep  

There are two distinct states of sleep; rapid-eye movement (REM) and non-rapid eye 

movement (NREM) sleep (Patel et al., 2021). These alternate cyclically, with each sleep cycle 

lasting about 90 minutes, although cycle duration varies across the sleep period and between 

individuals (Březinová, 1974). NREM sleep is further divided into 3 stages called N1, N2 and 

N3, based on brainwave characteristics (i.e. the electrical activity occurring in the brain), eye 

movements and muscle tone. Stage N1 is the lightest, and stage N3 is the deepest stage of 

sleep. Stage N3, also referred to as slow wave sleep (SWS) or deep sleep, is characterised by 

delta waves, or slow-wave activity. REM sleep is characterised by complete loss of muscle 



6 
 

tone, bursts of rapid eye movements, and dreaming (Carskadon & Dement, 2017). During REM 

sleep, brainwave activity looks similar to waking brainwave activity (Carskadon & Dement, 

2017).  

2.1.2 Measuring sleep  

Polysomnography (PSG) is considered to be the gold standard for measuring sleep (de Souza 

et al., 2003). PSG records cortical brain wave activity (electro-encephalogram), eye 

movements (electro-oculogram), and surface muscle tone (electromyogram), which are used 

to determine sleep onset latency, sleep efficiency, total sleep duration and sleep stages 

(McCall & McCall, 2012). PSG is often performed in a laboratory setting and is not practical for 

recording sleep across multiple days in field settings as it is expensive, time consuming and 

setting up the recording requires prior training (Arnal et al., 2020; Chinoy et al., 2021). 

Alternatively, actigraphy can be used to measure sleep, but unlike PSG, it cannot measure 

sleep states (McCall & McCall, 2012). Actigraphy is inexpensive and can easily be used for 

recording sleep-wake periods across several days-to-weeks (Ancoli-Israel et al., 2003; Ancoli-

Israel et al., 2015). An actigraph is placed on the non-dominant wrist and utilises an 

accelerometer to record activity. One minute is an epoch length commonly used, and when 

the device is set to record data in 1-minute epochs, movement is recorded as the number of 

activity counts per minute (i.e. wrist movement per epoch) (Ancoli-Israel et al., 2015). Sleep-

wake cycles are then identified using a sleep-scoring algorithm which has been validated 

against PSG (Ancoli-Israel et al., 2015). The sensitivity of actigraphy (i.e. the ability of 

actigraphy to detect sleep when PSG did the same) is reported to be high and satisfactory 

(Marino et al., 2013). However, the specificity of actigraphy (i.e. the ability of actigraphy to 

correctly identify wakefulness) is reported to be low (de Souza et al., 2003; Sadeh, 2011). 

Actigraphy can overestimate total sleep time and efficiency, and underestimate sleep onset 

latency (Paquet et al., 2007). This overestimation is due to the limited capacity of actigraphy 

to detect wakefulness when an individual is immobile in bed, but not asleep (de Souza et al., 

2003). Despite this, actigraphy has acceptable validity and reliability in healthy individuals with 

relatively good sleep patterns (Sadeh, 2011). 

2.1.3 Sleep regulation 

Two main processes have been proposed to regulate sleep: 1) the homeostatic drive for sleep 

(process S) and 2) the circadian drive (process C) (Borbély et al., 2016), as illustrated in Figure 



7 
 

2.1. The longer an individual is awake, the more slow‐wave activity will be present in the next 

sleep period, suggesting that slow wave activity reflects the brain’s need for sleep (Carskadon 

& Dement, 2017). Brain energy stores also predominantly replenish during SWS (Krueger et 

al., 2016). The homeostatic drive for sleep is thought to be triggered by the increased 

concentration of adenosine in the brain (Dworak et al., 2017; Reichert et al., 2022). Adenosine 

is a neurotransmitter produced by the increased breakdown of adenosine triphosphate (ATP) 

(Dworak et al., 2010). Adenosine inhibits the release of excitatory neurotransmitters, and, 

when adenosine accumulates during wakefulness, it decreases neuronal activity (Porkka-

Heiskanen & Kalinchuk, 2011). Adenosine may also influence the circadian clock functioning, 

but it still needs to be further investigated (Reichert et al., 2022). 

Circadian rhythms are daily biological cycles (the word ‘circadian’ comes from the Latin ‘circa 

diem’ = about a day), lasting approximately 24 hours (Duhart et al., 2021; Sharma & Kavuru, 

2010). Circadian rhythms are generated by the body’s internal master clock located within the 

suprachiasmatic nucleus (SCN) of the hypothalamus, and peripheral clocks located elsewhere 

in the brain and body (e.g. liver, lung, and skeletal muscle) (Yamanaka, 2020). They control the 

daily rhythm of sleep and wakefulness, and most biological processes (e.g. hormone 

production, metabolism, body temperature and cell regeneration) (Garbarino, 2020). In the 

absence of external time cues, the SCN produces rhythms with a period that, on average, is 

slightly longer than 24 hours. Due to a process of active synchronisation or entrainment, the 

SCN’s internal rhythm is able to keep pace with the external 24-hour light-dark cycle 

(Garbarino, 2020). Zeitgebers (meaning “time givers” in German) are time cues that regulate 

the SCN’s internal rhythm, with light being the most important one (Borbély et al., 2016). 

Sleep/wake, resting/active and feeding/fasting cycles also contribute to synchronising the SCN 

and circadian rhythms (Beersma & Gordijn, 2007; Duhart et al., 2021; Garbarino, 2020). Shift 

work, insufficient sunlight exposure, nocturnal exposure to artificial light and travel across 

multiple time zones, disrupt circadian rhythms and impair sleep (Garbarino et al., 2020). 

 



8 
 

 

Figure 2.1: The Two-Process Model of Sleep Regulation (adapted from Patanaik, 2015). 

This figure illustrates how process S (i.e. sleep homeostat) and process C (i.e. circadian process) 

interact. Whereas process C follows a pattern of an approximate 24-hour cycle, process S varies with 

sleep: it increases during the wake period and decreases during sleep. The difference between the 

two processes quantifies the sleep pressure. Sleep pressure increases with wakefulness and reduces 

with sleep (Borbély et al., 2016).  

2.1.4 Sleep restriction 

Sleep loss results from sleep restriction (SR) or total sleep deprivation (SD). To date, the 

majority of research has investigated SD as it generally induces large and consistent effects on 

cognition (Lim & Dinges, 2010). However, SR is experienced more frequently and by a greater 

proportion of the population (Banks et al., 2017; Rupp et al., 2012). Total SD is defined as being 

awake for a period beyond 16 to 18 hours (Banks et al., 2017), whereas SR is defined as not 

getting enough sleep per 24 hours for one or multiple nights (Banks et al., 2017). Insufficient 

sleep can be due to irregular work schedules, and long duty hours, personal or family 

commitments, social and sporting engagements, as well as individuals’ life choices or medical 

conditions (Banks et al., 2017; McLellan et al., 2016). 

Some individuals are more sensitive to sleep loss (Lowe et al., 2017), with resilience to sleep 

loss appearing to be a trait-like characteristic (Van Dongen et al., 2004), and cognitive domains 

affected by sleep loss may not be the same between individuals (Lowe et al., 2017). This 



9 
 

highlights the importance of controlling sleep duration prior to SR, or adjusting the magnitude 

of SR based on individuals prior sleep.  

2.1.4.1 Sleep restriction, cognition, and mood 

SR impairs cognitive functions and mood. In a recent meta-analysis by Lowe et al. (2017), SR 

was shown to compromise cognitive performance across multiple domains, particularly 

vigilance, reaction time, working memory and executive functions. The decline in cognitive 

performance following SR appears to have a dose-response effect (van Dongen et al., 2003). 

For example, Belenky et al. (2003) demonstrated that 7 days of either 3, 5, 7 hours TIB 

impaired 10-minute PVT performance (mean reciprocal response time; RRT, mean fastest 10% 

response time; RT, number of lapses) and increased subjective sleepiness compared with 9 

hours TIB. In the 5 and 7 hours TIB conditions, PVT performance decreased for the initial 5 

days before plateauing, whereas in the 3 hours TIB condition, PVT performance and subjective 

sleepiness continued to deteriorate (i.e. no stabilisation); however, sleepiness in the 5 and 7 

hours TIB conditions did not increase (Belenky et al., 2003). Similarly, van Dongen et al. (2003) 

demonstrated that 4 and 6 hours TIB per night for 14 consecutive days cumulatively impaired 

10-minute PVT performance (i.e. mean RRT, mean fastest 10% RT, number of lapses) and 

working memory measured using a computerised Digit Symbol Substitution Task relative to 8 

hours TIB. Similarly to Belenky et al. (2003), subjective sleepiness increased markedly during 

the initial 2 days, but only exhibited minor increases across subsequent days. This suggests 

that increases in subjective sleepiness does not always reciprocate reductions in cognitive 

performance following sleep restriction.  

It appears that a single night of SR is enough to affect cognition. Miyata et al. (2010) 

demonstrated that 4 hours TIB impaired reaction time and vigilance (assessed using a 

continuous performance test). Banks et al. (2010) also reported impaired performance on a 

10-minute PVT (i.e. number of lapses and mean fastest 10% RT) and 3-minute Digit Symbol 

Substitution Task (i.e. number of correct responses), and increased sleepiness following the 

first of five nights of SR (4 hours TIB). These performance effects typically coincide with 

impaired mood. For example, Baum et al (2014) showed that mood and vigour (assessed using 

the Profile of Mood States questionnaire) deteriorated following a few days of sleep 

restriction (6.5 hours TIB for 5 nights). Additionally, Kahn et al. (2014) reported higher fatigue 

and reduced vigour (assessed using the Profile of Mood States questionnaire), and increased 



10 
 

omission and commission errors in an online continuous performance test following a single 

night of SR (4 hours TIB). Overall, a single night of SR is likely to impair multiple aspects of 

cognition and mood.  

2.1.4.2 Vigilance, reaction time and Psychomotor Vigilance Task  

Vigilance, or sustained attention, is the “ability of an individual to maintain concentrated 

attention on specific target stimuli over a prolonged period” (Kim et al., 2017, p. 343). 

Vigilance is sensitive to the effects of SR (Lim & Dinges, 2010), and is commonly measured 

using the PVT (Dinges & Powell, 1985). The PVT is a popular, valid and reliable measure of 

cognitive performance during SR and total SD, and is sensitive to both circadian and 

homeostatic influences (Lim & Dinges, 2008, 2010). The PVT is simple to use and has minimal 

practice effects as performance does not improve by repeating the test, making it a preferred 

method for measuring vigilance in many laboratory and field-based studies (Basner et al., 

2018). The PVT requires a response to a visual stimulus that appears at random intervals, 

between 2 to 10 seconds, for a 10-minute test (Dinges & Powell, 1985). Shorter test durations 

often have a smaller inter-stimulus interval (e.g. 2-5 seconds) (Basner et al., 2018). The PVT 

can be easily administered via a computerised, handheld device and the most frequent 

measurements include reaction speed (reciprocal response time), mean 10% fastest and 

slowest reaction times and number of lapses (e.g. reaction times >500 ms).  

2.1.4.3 Sleep loss, brain energy metabolism and high-energy phosphates 

The brain is a highly metabolic organ, consuming ~20-25% of the resting metabolic rate in 

humans (Bailey, 2019). It is able to utilise various substrates for adenosine triphosphate (ATP) 

production (i.e. the energy currency of the cell), in particular glucose and, to a lesser extent, 

lactate, ketones and phosphorylcreatine (PCr). However, sleep loss disrupts brain energy 

metabolism, which may underpin impairments to cognitive functions. For example, during SD, 

the cerebral metabolic rate of glucose is reduced (Thomas et al., 2000) and glycogen (i.e. 

glucose polymer) may also decline (Kong et al., 2002). Recent interest, however, has focussed 

on the decline of PCr following sleep loss (Plante et al., 2014; Dowark et al., 2010), which is 

important as PCr acts by donating a phosphate group to adenosine diphosphate (ADP) to 

regenerate ATP following the hydrolysis of ATP to generate energy, ADP, and an inorganic 

phosphate (Pi). For example, Plante et al. (2014) sleep deprived 8 healthy adults (mean age 35 



11 
 

years old) for 40 hours and demonstrated with use of 31P Magnetic Resonance Spectroscopy 

that PCr declined (albeit non-significantly) by 3.9%, then increased by 7.3% after the first night 

of recovery sleep, and increased by a further 10.1% after the second night of recovery sleep. 

These findings were supported by Gordji-Nejadet al. (2018), who demonstrated, also with use 

of 31P Magnetic Resonance Spectroscopy, that cerebral PCr decreased during the day, but was 

restored after a 20-minute nap in 30 healthy adults (mean age 36 years old). Interestingly, in 

the same study, there was no significant change in cerebral ATP levels (Gordji-Nejad et al., 

2018). In rodents, Dworak et al. (2010) demonstrated that PCr decreased by 57% in the frontal 

cortex after 3 hours of SD while creatine levels tended to increase compared to undisturbed 

time matched controls. Together, these findings indicate that PCr is depleted to maintain ATP 

levels during extended wakefulness.  

2.1.5 Summary 

SR is common within the population and its detrimental effects accumulate with increased 

sleep loss. SR increases the risk of mishaps and accidents, and in catastrophic situations can 

lead to loss of life. Cognitive deficits related to SR are potentially influenced by disruptions to 

brain energy metabolism. Sleep loss appears to reduce PCr levels, which may limit the 

resynthesis of ATP, particularly during fatiguing cognitive tasks. Therefore, manipulating brain 

PCr stores prior to SR may alter subsequent energy metabolism and cognitive function. This 

can be achieved via orally ingesting supplemental creatine. 

2.2 What is creatine? 

Creatine is a naturally occurring guanidine compound involved in storage and transport of 

energy (Figure 2.2) (Walker, 1979). It supports high-energy phosphate metabolism, which is 

important for maintaining homeostasis of ATP and appears to act as a cellular energy buffer 

(Walker, 1979). During periods of accelerated ATP degradation, such as when performing a 

cognitive-demanding task, creatine facilitates rapid energy provision by transferring a 

phosphate from PCr to ADP to regenerate ATP (Greenhaff, 2001; Walker, 1979). This reaction 

is reversibly catalysed by the enzyme, creatine kinase (CK) (Greenhaff, 2001; Walker, 1979). 

This energy pathway is typically referred to as the PCr/CK/ATP system and has been named 

the Lohman reaction (Figure 2.3). At energy producing sites in the cell, creatine is 

phosphorylated by CK to PCr, which joins the cell’s PCr pool and is used to phosphyorylate 

adenine compounds, primarily ADP. PCr serves as a rapidly available reserve of high-energy 



12 
 

phosphates in the brain to recycle ATP. The generated ATP contributes to energy-required 

within the cell.  

Figure 2.2: Creatine molecule. 

Creatine is found mostly in muscle tissue, hence its name which derives from kreas, the Greek 

word for meat (Rae & Bröer, 2015). Creatine is also stored in other tissues, such as the heart 

and the brain (Jacobus & Diffley, 1986; Monge et al., 2008). Creatine supplements have been 

largely used by athletes and weightlifters to increase muscle creatine stores and enhance 

physical performance, particularly for explosive and high-intensity exercise (Kreider, 2003). 

Relative to muscle, the brain contains less than 5% of the total body creatine (i.e. total of ~120 

g in a 70 kg human), but similarly, it is highly metabolically active and requires creatine for 

energy production (Gualano et al., 2010; Turner et al., 2015). This has led to research assessing 

how increasing brain creatine affects its function, particularly when under stress and in 

diseased states (Gualano et al., 2010; Turner et al., 2015). For example, creatine 

supplementation has been shown to mitigate symptoms (e.g. decline in vigilance) in situations 

where neural energy provision is impaired such as hypoxia (Turner et al., 2015). It is also 

possible that increasing brain creatine could be beneficial during other stressors that disrupt 

metabolic energy production, such as sleep loss.  

 

Figure 2.3: Lohman reaction (adapted from Turner, 2015). 

 



13 
 

2.2.1 Sources of creatine 

Humans require ~2 gday-1 of creatine to maintain creatine stores (Wyss & Kaddurah-Daouk, 

2000). Approximately 1 g can be obtained via an omnivorous diet including meat and fish 

(Andres et al., 2008; Wallimann et al., 2007), and the other half is via endogenous biosynthesis 

(East, 2002), which is an inter-organ process involving the kidneys, pancreas, and liver (Walker, 

1979). The brain can also synthesise creatine and does not necessarily rely on dietary or 

endogenous production from other organs (Braissant et al., 2007). 

Endogenous synthesis of creatine is a two-step reaction and requires three amino acids: 

glycine, arginine, and methionine (Brosnan et al., 2011). Initially, the amino acids arginine and 

glycine are combined within the kidneys by an enzyme (L-arginine-glycine amidino 

transferase) to form guanidinoacetate, and the amino acid ornithine is released as a secondary 

product. Guadinoacetate travels to the liver and is methylated by S-adenosyl-L-methionine in 

a reaction catalysed by the enzyme guanodinoacetate- methyltransferase. This produces 

creatine and S-adenosylhomocysteine (Figure 2.4). Creatine is then predominantly 

transported to the skeletal muscle and the brain. Once in the cells, creatine can be 

phosphorylated to PCr in a Lohman reaction (Figure 2.3). The phosphate group comes from 

ATP (Wyss & Kaddurah-Daouk, 2000).  

 

Glycine + arginine  Guanidinoacetate + Ornithine           Adenosylhomocysteine + Creatine 

 

 

Figure 2.4: Creatine synthesis. 

Creatine biosynthesis and absorption are regulated by dietary creatine sources (Lukaszuk et 

al., 2002). Individuals frequently ingesting creatine via meat or fish have lower rates of 

endogenous creatine synthesis than individuals who ingest little or no dietary creatine (i.e. 

vegetarians and vegans) (Venderley & Campbell, 2006). In addition to dietary and endogenous 

sources, creatine can also be ingested in supplemental form. Creatine supplements are 

generally safe, cheap, and easily accessible, and are often used to increase tissue creatine 

content (Kim et al., 2011; Persky & Rawson, 2007; Yoshizumi & Tsourounis, 2004).  

S-Adenosylmethionine 



14 
 

2.2.2 Optimal protocol for creatine supplementation 

Creatine supplements are available in various forms; however, creatine monohydrate (CrM) is 

the most commonly used and will be the focus of this review. Similar to the muscle, brain 

creatine stores can be increased via creatine supplementation (Dechent et al., 1999; Lyoo et 

al., 2003; Turner, 2015; Turner et al., 2015). Unlike the muscle, however, optimal creatine 

supplementation protocols to increase brain creatine levels are less clear (Roschel et al., 

2021). Whilst short term creatine supplementation (e.g. 20 gday-1 for 5-7 days) can increase 

intramuscular creatine stores by ~10% (Becque et al., 2000) to 40% (Kreider, 2003), it appears 

creatine supplementation can only increase brain creatine and PCr by ~5-10% (Dechent et al., 

1999; Turner et al., 2015). For example, in 15 individuals (mean age 31 years old) ingesting 20 

gday-1 of CrM split evenly into 4 daily 5 g doses for 7 days, brain creatine increased by 9.2% 

on average (Turner et al., 2015). A similar dose given for 4 weeks in 6 healthy participants 

(mean age 26.8 years old) has also shown to increase total brain creatine by 8.7% (Dechent et 

al., 1999). As such, the brain appears less sensitive to exogenous creatine sources than 

muscles, which may be due to mechanisms preventing orally ingested creatine from being 

transported into brain cells (Dolan et al., 2019; Roschel et al., 2021). Therefore, high doses (i.e. 

~20 g) of creatine across several days are likely needed to increase brain creatine and PCr 

(Roschel et al., 2021).  

2.2.3 Safety of creatine supplementation 

CrM is considered safe to ingest within healthy individuals (European Food Safety Acency, 

2004). Some possible side effects include a slight weight gain caused by increased storage of 

water, muscle and gastrointestinal cramps (Juhn, 1999; Vandenberghe et al., 1997); however, 

side effects often appear anecdotal and are likely to be due to inappropriate creatine dosing 

protocols (e.g. >10 g for a single dose for 28 days) (Ostojic & Ahmetovic, 2008), or may not be 

specifically related to creatine supplementation (e.g. muscle cramps can result from 

dehydration) (Dalbo et al., 2008; Kim et al., 2011). Additionally, there is no evidence that 

creatine supplementation adversely affects the kidneys in healthy individuals (Hall & Trojian, 

2013). Altogether, CrM may be used without concern by healthy individuals when ingested in 

recommended quantities.  



15 
 

2.2.4 Creatine supplementation and cognition 

The effect of creatine supplementation on cognition at rest (i.e. under no physiological stress) 

appears equivocal. Several studies have demonstrated no effect (Benton & Donohoe, 2011; 

Rae et al., 2003; Rawson et al., 2008; van Cutsem et al., 2020; Watanabe et al., 2002), but a 

selected few have demonstrated beneficial effects (Rae et al., 2003; van Cutsem et al., 2020; 

Watanabe et al., 2002). These favourable effects may have been due to the population being 

vegetarians (Rae & Bröer, 2015), who typically have lower creatine stores and are more 

sensitive to creatine supplementation (Solis et al., 2017), or because the cognitive tasks were 

prolonged and fatiguing (van Cutsem et al., 2020; Watanabe et al., 2002), thus stressing brain 

energy metabolism. For example, Watanabe et al. (2002) demonstrated that 8 gday-1 CrM for 

5 days increased the number of correct responses on a 15-minute serial calculation task in 24 

healthy individuals (mean age 24.3 years old). Moreover, van Cutsem et al. (2020) 

demonstrated that 20 gday-1 CrM for 7 days improved accuracy during a 90-minute Stroop 

task; however, it did not alter short cognitive performance (Flanker task) and vigour. 

Furthermore, when concurrently experiencing a physiological stressor, the benefits of 

creatine supplementation may be more pronounced. For example, 20 gday-1 CrM for 7 days 

improved multiple aspects of cognition during exposure to hypoxia (10% oxygen) in 15 healthy 

individuals (mean age 31 years old) (Turner et al., 2015). Altogether, it appears that creatine 

supplementation may be beneficial for improving performance on sufficiently fatiguing 

cognitive tasks, and that these benefits may more likely manifest when cerebral energy 

metabolism is stressed, such as following SR.  

2.2.5 Creatine supplementation, sleep loss and cognition 

Creatine supplementation may influence cognition following sleep loss. To date, the effect of 

creatine supplementation on cognition in humans following sleep loss has been investigated 

in three studies (Cook et al., 2011; McMorris et al., 2007; McMorris et al., 2006) (Table 2.1). 

McMorris et al. (2006) supplemented 10 healthy adults with 20 gday-1 of CrM for 7 days and, 

when compared to the placebo group (n = 9), found no cognitive performance or mood effects 

after 6 and 12 hours of SD; however, after 24 hours of SD, changes in working memory 

(measured via random movement generation test), choice reaction time, and mood were 

smaller in the creatine supplemented group than in the placebo group. In another study, 

McMorris et al. (2007) also supplemented 10 healthy adults with 20 gday-1 of CrM for 7 days 



16 
 

(placebo, n = 9) and demonstrated that after 36 hours of SD, but not 18 or 24 hours, 

performance on a central executive task was less degraded compared to the placebo group. 

In that same study, there were no differences between groups for mood, choice reaction time 

or working memory (McMorris et al., 2007). Collectively, these findings suggest that creatine 

loading can mitigate degradation in aspects of cognition following sufficient sleep loss, but 

these effects seem equivocal.  

Creatine supplementation may also be able to alter cognition following SR. In a study by Cook 

et al. (2011), 10 elite rugby players were supplemented with an acute dose of CrM of 50 to 

100 mgkg-1 1.5 hours prior to repeated performance on a passing task. Interestingly, pass 

accuracy was less impaired following a single night of 3-5 hours of sleep (Cook et al., 2011), 

suggesting CrM supplementation had a positive effect on psychomotor skills. However, effects 

on cognitive performance and mood were not assessed. The mechanism of how CrM could 

have positively effect skill performance however is uncertain, as there is no evidence 

regarding the effect of a single acute dose of creatine on cerebral energy metabolism. In rats, 

SD induced-increase in extracellular adenosine was attenuated after CrM supplementation (a 

rodent diet supplemented with 2% of CrM for 4 weeks). This finding suggests that CrM 

supplementation may also be able to reduce sleep need and homeostatic sleep pressure 

following SD (Dworak et al., 2017), although these effects have not been replicated in humans. 

Further research is required to assess if creatine supplementation can alter cognitive 

performance following SR and whether it also influences other aspects of cognition, such as 

sleepiness and mood.  



17 
 

Table 2.1: The effect of creatine supplementation on cognition following sleep loss.  

Author Population Study design SR/SD protocol  
Creatine 

supplementation 
protocol 

Cognitive and/or 
performance tests 

Primary outcomes 

McMorris 
et al. 
(2006) 

n = 19, mean 
age = 21.1, 
young 
athletic male 
and female  

Double-
blinded, 
placebo-
controlled, 
parallel design 

Series of tests at 
baseline and after 
6, 12 and 24 h of 
SD w/ moderate 
intermittent 
exercise after a 7-
day CrM loading or 
placebo intake 
 

5 g 4 times per 
day for 7 days  

Random movement 
generation and adjacency, 
verbal and spatial recall, 
choice reaction time, static 
balance, and mood state  

Performance reduction was 
attenuated in the CrM group for 
random movement generation, 
choice reaction time, balance, 
and mood following 24 h of SD 
 
 

McMorris 
et al. 
(2007) 

n = 19, mean 
age = 21.1, 
young 
athletic male  

Double 
blinded, 
placebo-
controlled, 
parallel design  

Series of tests at 
baseline and after 
18, 24 and 36 h of 
SD w/ moderate 
intermittent 
exercise after a 7-
day CrM loading or 
placebo intake 
 

5 g 4 times per 
day for 7 days  

Random number 
generation test, verbal 
short memory test, choice 
reaction time, dynamic 
balance test, shortened 
version of the Profile of 
Mood States 
questionnaire, NASA-TLX 

CrM supplementation only 
affects the random number 
generation test performance 
following 36 h of SD.  
CrM had no effect on mood 
state 

Cook et al. 
(2011) 

n = 10, mean 
age = 20, 
elite male 
rugby players  

Blinded, 
placebo-
controlled 
crossover trial 

Complete 5 trials 
on a simple rugby 
passing skill test 
after 7-9 h of sleep 
and 5 trials after 3-
5 h of sleep  

1 or 5 mgkg-1 1 
of caffeine, 50 to 

100 mgkg-1 CrM 
1.5 h before each 
trial 

Simple rugby passing skill 
test (20 repeats per trial) 

SD reduced passing accuracy 
and CrM supplementation 
reversed this effect (trend for 
greater effect in larger dose) 

  



18 
 

2.3 Conclusion 

The brain is a highly metabolically active organ and relies on a constant energy supply. During 

times of accelerated brain ATP turnover, such as fatiguing cognitive tasks, creatine 

supplementation over several days can facilitate a more constant regeneration of ATP in the 

brain. Therefore, prophylactic creatine supplementation may protect brain function in the 

event of acute stressors, such as SR. Despite promising research demonstrating some benefit 

of creatine supplementation on cognition following total SD, there is currently no evidence in 

the context of SR, except for a single study assessing psychomotor performance. Further 

research is clearly required.  

2.3.1 Comment from the author  

Chapter 3 contains a subset of data from the main study as time restrictions, and the scope of 

this thesis, did not permit the inclusion of all data collected. The additional variables 

measured, but not reported on, are listed in the ‘Project Procedures’ section of the Participant 

Information Sheet, which can be found in Appendix A.  

 

  



19 
 

3 Research Manuscript 

Title 

The effect of creatine monohydrate loading on cognitive performance, mood, sleepiness, and 

perceived workload following sleep restriction. 

Authors 

Janvresse, Juliette1 

van den Berg, Margo2 

Bloomfield, Peter3 

Gant, Nick3 

Shaw, David M1,4 

Institutions 

1School of Sport, Exercise and Nutrition, Massey University, New Zealand  

2Sleep/Wake Research Centre, School of Health Sciences, Massey University, New Zealand 

3Department of Exercise Sciences, University of Auckland, New Zealand 

4Aviation Medicine Unit, Royal New Zealand Air Force Base Auckland, Whenuapai, Auckland, 

New Zealand  



20 
 

3.1 Abstract 

Background: Sleep restriction (SR; i.e. not getting enough sleep per 24 hours for one or 

multiple nights) impairs cognitive performance and mood, which can compromise capability 

and safety. SR-related performance deficits may be underpinned by disruptions to brain 

energy metabolism. Sleep loss appears to reduce brain phosphorylcreatine (PCr; also called 

creatine phosphate or phosphocreatine), which supports the regeneration of adenosine 

triphosphate (ATP) from adenosine diphosphate (ADP). Since creatine supplementation 

increases tissue PCr, including in the brain, it may mitigate possible SR-related cognitive 

impairments.  

Aim: To assess the effect of a 7-day creatine monohydrate (CrM) loading protocol on cognitive 

performance, mood, subjective sleepiness, and perceived workload following one night of SR 

(3 hours time-in-bed; TIB). 

Methods: A double-blinded, randomised, placebo-controlled cross-over trial was conducted 

with 7 healthy participants (3 men and 4 women); age, 28.0 ± 4.6 years (range, 22-35 years). 

Participants ingested 20 gday-1 of CrM or placebo (i.e. tapioca) divided into 4 equal doses for 

7 days followed by a single SR night (3 hours TIB). Cognitive performance (10-minute 

Psychomotor Vigilance Task; PVT), mood (fatigue and vigour), subjective sleepiness, and 

perceived workload were measured following ~8 hours of sleep prior to creatine 

supplementation and following the CrM loading protocol and SR. A minimum of a 5-week 

wash-out period separated the two trial arms. Linear mixed effects models were employed to 

analyse data.  

Results: SR impaired mean reciprocal reaction time (p = 0.01) and mean 10% fastest reaction 

time (p = 0.02), increased fatigue (p = 0.004), subjective sleepiness (p = 0.009), and workload 

(p = 0.02), and reduced vigour (p <0.001); however, it did not alter the number of lapses (p = 

0.60) or mean 10% slowest reaction time (p = 0.60). CrM loading had no effect on cognitive 

performance, mood, subjective sleepiness, or workload. 

Conclusion: CrM loading may not alter the detrimental effects of a single night of SR (3 hours 

TIB) on 10-minute PVT performance, mood, subjective sleepiness, or perceived workload.  

Key words: sleep loss; psychomotor vigilance task; randomised cross-over trial 



21 
 

3.2 Introduction 

Sleep is essential for optimal health and performance (Hirshkowitz et al., 2015; Watson et al., 

2015). Whilst the recommended sleep duration for adult is 7 to 9 hours per night (Hirshkowitz 

et al., 2015; Watson et al., 2015), SR (i.e. not getting enough sleep per 24 hours for one or 

multiple nights) is common (Banks et al., 2017; McLellan et al., 2016; Rupp et al., 2012). SR 

can result from work schedules overlapping with the person’s normal sleep time, long duty 

hours, personal or family commitments, or medical conditions (Banks et al., 2017; McLellan et 

al., 2016; Rupp et al., 2012). SR increases daytime fatigue and impairs cognitive performance 

and mood (Banks et al., 2017; Lowe et al., 2017), resulting in increased risk of incidents and 

accidents (Caldwell et al., 2019). Caffeine consumption remains the main dietary strategy to 

mitigate these effects (Irwin et al., 2020); however, high caffeine intakes can exacerbate SR by 

negatively affecting sleep quality and sleep duration (Temple et al., 2018). Alternative 

strategies should therefore be considered, such as creatine supplementation.  

Creatine is derived from dietary sources (~50%), mostly meat and fish, and via endogenous 

synthesis in the kidneys and liver (Andres et al., 2008; East, 2002; Wallimann et al., 2007). In 

order to increase tissue creatine content, supplementation is typically required. For example, 

20 gday-1 of CrM for 7 days increased brain creatine by 9.2% (Turner et al., 2015). CrM 

supplementation is safe (European Food Safety Acency, 2004; Kim et al., 2011) and is typically 

ingested over several days-to-weeks (Lyoo et al., 2012). Several studies have assessed the 

effect of creatine supplementation on cognition (Avgerinos et al., 2018), but few have 

demonstrated beneficial effects (Rae et al., 2003; van Cutsem et al., 2020; Watanabe et al., 

2002). Creatine supplementation also appears more beneficial for cognitive performance 

during physiological stress, such as hypoxia (Turner, 2015). Considering sleep loss appears to 

reduce brain PCr concentration (Plante et al., 2014), which is important for the resynthesis of 

the brain's primary energy source (i.e. ATP), creatine supplementation may be able to support 

cognition following SR.  

Indeed, creatine supplementation has demonstrated beneficial effects on cognition following 

total sleep deprivation (SD) (McMorris et al., 2007; McMorris et al., 2006). Considering SR is 

more common than SD and creatine supplementation effects may differ when sleep is 

attained but in insufficient amounts, further research is required specifically in the context of 

SR. Therefore, the aim of this research was to assess the effect of a 7-day creatine loading 



22 
 

protocol on simple cognitive task performance, mood, subjective sleepiness, and perceived 

workload following one night of SR (3 hours TIB). Findings will contribute to determining 

whether creatine can be ingested as a countermeasure to sleep loss related performance 

impairment. 

3.3 Material and Methods 

3.3.1 Study overview 

A double-blinded, randomised, placebo-controlled, cross-over design trial was conducted in 

healthy men and women to compare the effect of CrM loading and a placebo (PLA) on 

cognitive performance, mood, subjective sleepiness, and perceived workload following SR (3 

hours TIB). Ethical approval was provided by the Massey University Ethics Committee (SOA 

21/47). Participants were fully informed of the study’s rationale and possible risks of the 

experimental procedures before providing their written consent.  

3.3.2 Eligibility criteria 

Participants were recruited via emails, posters dispatched at Massey University and word-of-

mouth. Participants were required to be: 1) healthy; 2) aged 18-50 years; 3) body mass index 

<30 kgm²; 4) consuming a mixed diet; 5) habitually going to bed between 2100 and 0000 and 

waking between 0600 and 0900. In addition, women were eligible to participate ensuring both 

trial arms were during the same periods of their menstrual cycle. Participants were excluded 

if they: 1) supplemented with creatine in the previous 3 months; 2) smoked; 3) consumed >3 

cups caffeinated beverages per day; 4) habitually slept <7 hours or >9 hours per night; 5) 

regularly consumed medications acting on the central nervous system; 6) had a history of drug 

or alcohol abuse; 7) engaged in shift work or trans-meridian travel within the 28 days prior to 

the study. Participants were also required to have: Epworth Sleepiness Scale score <10 (Johns, 

1991); global Pittsburg Sleep Quality Index score ≤5 (Buysse, Reynolds, Monk, Berman, 

&Kupfer, 1989); normal scores on the 21-item Depression Anxiety Stress Scale (i.e. 0-7) 

(Lovibond & Lovibond, 1995); and scoring as either moderately evening or intermediate 

chronotype on the Horne-Östberg Morningness/Eveningness Questionnaire (Horne & 

Östberg, 1976). Participants were fully informed of the rationale of the study and possible 

risks of the experimental procedures (Appendix A) before providing their written consent; 



23 
 

however, they were not informed of the potential cognitive effects of a CrM loading. Each 

participant received a calendar summarizing their daily task (Appendix B). 

3.3.3 Experimental procedures 

Figure 3.1. provides an overview of the experimental design. Prior to the study, participants 

were familiarised with the cognitive test battery by performing it 5 times to minimise practice 

effects. In the morning following two nights of at least 8 hours TIB (approximately 2200-2230 

to 0600-0630), participants presented to the laboratory between 0730 and 0930 in a fasted 

and rested state to complete the testing protocol (day 1; baseline). Immediately afterwards, 

participants commenced their dietary supplement allocation (i.e. CrM or PLA) for 7 days as 

per prior randomisation (www.randomizer.org). Diet records were reported on 2 non-

consecutive days and sleep was monitored daily via wrist actigraphy. Participants were 

requested to maintain habitual bed and rise times (approximately 8 hours TIB) between days 

1 to 6. On day 7, participants were requested to restrict sleep to 3 hours TIB (0300 to 0600). 

Regular text messages on WhatsApp and actigraphy data confirmed participants respected 

the protocol. During the study, participants were requested to abstain from over-the-counter 

medications, alcohol, napping, and remain hydrated. Caffeine intake was limited to <100 

mgday-1 ingested before 1200 and prohibited for the 24 hours prior to testing (Appendix C). 

Ingestion of meat and fish was prohibited during the 24 hours prior to testing. Physical activity 

was requested to be maintained at moderate levels. All other lifestyle choices were allowed 

to vary naturally during the study period. Following SR (day 8; post-intervention), participants 

presented to the laboratory between 0730 and 0930 in a fasted state to repeat the testing 

protocol. In order to avoid participants driving themselves following SR and risk a fatigue-

related accident, participants’ transport to and from the laboratory was arranged and paid 

for. Following completion, participants were asked to state what supplement protocol they 

were randomised to. The washout period between treatments was at least 5 weeks as this 

appears to reduce tissue creatine concentrations to approximately baseline (pre-supplement) 

levels (Hultman et al., 1996; McKenna et al., 1999; Preen et al., 2003). After the washout 

period, participants repeated the experimental protocol with the alternative supplement 

treatment.  



24 
 

3.3.4 Dietary supplementation 

Participants ingested 20 gday-1 of CrM (NZ Muscle, New Zealand) or tapioca (Pams, New 

Zealand). Supplements were made in a research grade kitchen and provided in powdered form 

mixed with 3 g of a flavoured drink powder (Vitafresh, New Zealand) to be mixed with water 

and taste matched for blinding. Participants were requested to ingest 4 CrM doses of 5 g each 

daily at equally spaced intervals on days 1 to 7. A CrM loading strategy of 20 gday-1 for 7 days 

was chosen as this has been shown to augment creatine stores within the brain of men and 

women (Turner et al., 2015). Participants were requested to send a text message to the 

researchers to confirm ingestion at the end of each day.  

 

 

Figure 3.1: An overview of the experimental design. Abbreviations: TIB, time-in-bed; DS, 
dietary supplementation. 

3.3.5 Testing protocol 

Following arrival at Massey University (Albany Campus), participants rested in a quiet room 

for 10 min, then performed a 10-minute PVT. Immediately afterwards, mood and sleepiness 

were measured using the Automated Neuropsychological Assessment Metrics test battery 

(ANAM®) and subjective workload on a written handout. All testing was performed at the same 

time of day, in a room under fluorescent lighting and constant temperature (~20 oC). 

3.3.5.1 PVT-192 task 

Vigilant attention was measured using a 10-minute version of the PVT (Dinges & Powell, 1985). 

Participants respond to the presentation of a visual stimulus as quickly as possible by pressing 

a button with the thumb of their dominant hand (Dinges & Powell, 1985). After a random time 

interval (between 2 to 10 seconds), the stimulus reappears. The PVT is included in this study 



25 
 

because of its sensitivity to the effects of fatigue and short administration time (Lim & Dinges, 

2008, 2010). The PVT is also very reliable with little evidence of practice effects (Balkin et al., 

2004). Variables for analysis included number of lapses (i.e. reaction times exceeding 500 ms), 

mean reciprocal response time (RRT; 1/mean response time (RT)] × 1000), mean fastest 10% 

RT and mean slowest 10% RT. 

3.3.5.2 Mood 

Fatigue and vigour were measured using the computerised ANAM® mood scale. Validated 

against the Profile of Mood Scores (POMS) (Johnson, Vincent, Johnson, Gilliland, & Schlegel, 

2008), each variable consists of six adjectives displayed on a 7-point Likert scale anchored with 

0 (not at all) to 6 (very much). Fatigue adjectives included lazy, inactive, tired, weary, sluggish, 

and drowsy, whilst vigour adjectives included energetic, lively, alert, spirited, active, and 

vigorous. Higher scores indicated more fatigue or vigour. 

3.3.5.3 Subjective sleepiness 

Subjective sleepiness was measured in the ANAM test battery using the Stanford Sleepiness 

Scale (Hoddes, Zarcone, Smythe, Phillips, & Dement, 1973). Participants were asked to rate 

their sleepiness on a 7-point Likert scale which was anchored from 1 (feeling active, vital, alert, 

or wide awake) to 7 (cannot stay awake, sleep onset appears imminent). Higher scores 

indicated greater perceived sleepiness. 

3.3.5.4 Subjective workload 

Subjective workload was measured using the modified National Aeronautics and Space 

Administration Task Load Index (NASA-TLX) (G.Hart & E.Staveland, 1988), whereby all 

subscales have an equal weight (i.e. Raw-TLX) (Hart, 2006). Each bipolar sub-scale (mental 

demand, physical demand, temporal demand, performance, effort, and frustration) is 

anchored with the verbal descriptions ‘Low’ to ‘High’ (except for the performance sub-scale, 

which is anchored with the verbal descriptions ‘Good’ to ‘Poor’) and divided into twenty equal 

intervals. Participants were instructed to place an ‘X’ in the position that best describes their 

experience from 0 to 100. Overall workload scores were calculated through averaging the sum 

of the subscales, with a higher mean score indicating higher perceived workload. 



26 
 

3.3.6 Dietary monitoring 

Participants were asked to complete two diet records, one prior intervention and one during 

the dietary supplementation. Diet records were ascertained using image-assisted (alongside a 

fiducial marker), weighed diet records reported in real-time to a New Zealand registered 

dietitian via a mobile phone application (WhatsApp, Facebook, San Francisco, CA). Diet 

records were assessed via visual inspection.  

3.3.7 Sleep monitoring 

Sleep was monitored using a wrist actigraph (Micro Motionlogger Watch; Ambulatory 

Monitoring Inc., Ardsley, New York, USA), worn on the participant’s non-dominant wrist. 

Participants were instructed to press the event marker on the actigraph to indicate when they 

began trying to sleep and when they finished trying to sleep. A standardised sleep diary was 

also completed to support actigraph data. Actigraphy objectively assesses sleep-wake 

patterns and has been validated against the gold standard polysomnography (Ancoli-Israel et 

al., 2015). Data was recorded in 1-minute epochs and analysed by an experienced sleep 

researcher using ActionW2.7 software with the Cole-Kripke sleep-scoring algorithm applied, 

in conjunction with information from the sleep log. The variable chosen for analysis was sleep 

duration, defined as the total number of minutes scored as sleep within the sleep interval 

(excluding any minutes scored as wake) (Cole et al., 1992). Participants were asked to aim for 

7 to 9 hours of sleep per night and  restrict sleep on the 7th night of the intervention (3 hours 

TIB; 0300 to 0600). A text message requesting a response was sent to participants on the 

morning following the SR night to ensure participants were awake at 0600.  

3.3.8 Statistical Analysis 

All statistical analyses were conducted in R version 3.6.0 with RStudio version 1.1463 (R Core 

Team, 2020). Sleep duration was analysed using paired t-tests or Wilcoxon signed rank test 

for normally and non-normally distributed data, respectively; however, as a data point was 

missing for sleep duration in the supplementation period, a Welch test was employed. 

Normality was assess using a Shapiro-Wilk test. All other data were analysed using a series of 

linear mixed models with restricted maximum likelihood using the R package “lme4”. Fixed 

effects included condition (2 levels; creatine and placebo) and test (2 levels; baseline and sleep 

restricted). A random intercept for participant was included to adjust for inter-individual 



27 
 

variability and a random slope for sleep restriction was included to adjust for differences in 

the effect of sleep between participants. Models were controlled for supplement order. 

Normality of distribution and homoscedasticity of the model’s residuals were determined by 

visual inspection of Q-Q plots; if violated, data were either log, square-root or inverse 

transformed and assessed for best fit prior to extracting p-values. P-values for fixed-effects 

factors were obtained using Type II Wald F tests with Kenward-Roger degrees of freedom in 

the R package “car”. P-values for pairwise comparisons were obtained using Holm adjustment 

for multiplicity in the R package “emmeans”. Significance was inferred when p  0.05. Graphs 

were produced in GraphPad Prism Version 9.1.1 (GraphPad Software, San Diego, California 

USA).  

3.3.9 Comment from the author 

R scripts for data analyses were written by David Shaw; however, I executed each step and 

interpreted the outputs.  

 

3.4 Results 

3.4.1 Participants 

15 individuals volunteered for this study and were assessed for eligibility; however, 5 were 

either ineligible or decided not to participate. 10 participants commenced the study but 3 

were removed due to not adhering to the study’s requirements. A total of 7 participants (3 

men and 4 women); age, 28.0 ± 4.6 years (range, 22-35 years); body mass, 68.1 ± 14.0 kg; 

height, 1.70 m ± 0.12 m; body mass index, 23.1 ± 2.0 kg.m² were included in the data analyses 

(Figure 3.2). Two of the participants were part of the research team, but the double-blinding 

study setting ensured their responses were not biased. Due to the resulting low sample size; 

the study was treated as a pilot study.  

 

 

 

 



28 
 

  

 

 

 

 

 

 

 

 

 

 

 

Figure 3.2: Participant selection. 

3.4.2 Supplement and dietary compliance 

Participants reported consistent dietary habits during each supplement intervention, which 

was confirmed by a visual assessment of their 2-day weighed image-assisted dietary records. 

Participants also confirmed not consuming meat or fish on the day prior to performing the 

cognitive tests. Compliance to the supplementation protocols appeared to be excellent as 

confirmed by participants’ daily messages of ingestion to the researchers. 

 

 

Referrals assessed for 

eligibility (n = 15) 

Excluded: 

• Did not meet inclusion criteria (n = 2) 

• Declined to participate (n = 3) 

Recruited (n = 10) 

Completed familiarisation (n = 10) 

Excluded as slept >3 hours TIB (n = 3) 

Analysed (n = 7) 



29 
 

3.4.3 Sleep duration 

3.4.3.1 2-day baseline 

Average sleep duration on baseline days did not significantly differ between the placebo and 

CrM interventions (p = 0.973) (Table 3.1). All participants went to bed between 2100 and 0000 

at baseline for both placebo and CrM, except for participant 3 who went to sleep at 0036 on 

the first baseline night in the CrM intervention. Most participants obtained at least 7 hours 

sleep each night; however, participant 3 slept <7 hours on all baseline nights (6.15 - 6.83 

hours) and participant 11 obtained 5.31 hours sleep on the first baseline night in the placebo 

condition. Since sleep on baseline days was not expected to affect test results following SR 

(day 8), data from participants 3 and 11 remained included. 

3.4.3.2 7-day supplement intervention  

Average sleep duration on days 1 to 6 did not differ between placebo and CrM interventions 

(p = 0.393) (Table 3.1). All participants went to bed between 2100 and 0000 and slept at least 

7 hours each night during the 7-day CrM and placebo interventions. Participant 7 had a short 

nap on day 5 during the placebo intervention, but only obtained 1 minute sleep; since this was 

not expected to affect results, participant’s 7 data remained included. For the placebo 

intervention, participants 5, 7 and 11 slept >9 hours (9.17 - 10.48 hours) for three to four 

nights. For the CrM intervention, participants 1, 6 and 7 slept >9 hours for one night (9.18 - 

9.62 hours) and participant 5 slept >9 hours for four nights (9.32 - 10.02 hours). As participant 

6 did not wear his watch on day 7 during the placebo intervention, the sleep duration for this 

day was not available. During the CrM intervention, participants 11 and 6 had one split sleep 

(i.e. night-time sleep was divided in two parts) on days 2 and 8, and participant 3 had two split 

sleeps on days 3 and 7.  

3.4.3.3 Sleep restriction night 

Sleep duration on the SR night did not significantly differ between conditions (p = 0.435) (Table 

3.1). For the placebo intervention, participant 11 slept slightly more than 3 hours (i.e. 3.18 

hours). For the CrM intervention, participants 6 and 11 spent >3 hours TIB (i.e. 3.48 and 3.62 

hours), with participant 11 obtaining 3.03 hours sleep, while participant 6 obtained 1.98 hours 

of sleep. 



30 
 

Table 3.1: Sleep duration for baseline, 7-day intervention and SR for CrM and placebo (in 
hours). 

 Placebo CrM 
Study period Mean ± SD Median Range n Mean ± SD Median Range n 

Baseline Days 7.75±1.17 7.87 5.32- 
9.83 

14 
7.76±0.76 7.83 6.15- 

9.23 
14 

7-day 
Intervention 

7.86±0.93 7.93 6.10- 
9.92 

48a 
7.72± 0.70 7.82 5.90- 

9.53 
49 

SR Night 2.78±0.22 2.78 2.52- 
3.18 

7 
2.70±0.36 2.75 1.98- 

3.03 
7 

ᵃ Includes one outlier. 

 

3.4.3.4 Duration of wakefulness prior to the cognitive tests 

There were no differences between interventions for time awake prior cognitive tests at 

baseline (p = 0.501) or following SR (p = 0.578). Within each supplement condition, there were 

no differences between baseline between baseline and SR for time awake prior cognitive 

testing (placebo: p = 0.318; CrM: p = 0.219).  

Table 3.2: Duration of wakefulness prior to each testing time (in hours). 

 Placebo CrM 
Study period Mean ± SD Median Range n Mean ± SD Median Range n 

Baseline 1.54 ± 0.44 1.48 
1.02- 
2.08 

7 1.41 ± 0.61 1.22 
0.60- 
2.55 

7 

SR 1.72 ± 0.59 1.67 
0.98- 
2.48 

7 1.63ᵃ ± 0.33 1.53 
1.05- 
2.02 

7 

ᵃ Data not normally distributed. 

 

3.4.4 PVT performance 

SR significantly increased mean fastest 10% RT (p = 0.017; Figure 3.3B) and decreased mean 

RRT (p = 0.013; Figure 3.3D) but did not affect mean slowest 10% RT (p = 0.602, Figure 3.3A) 

or number of lapses (p = 0.599; Figure 3.3C). CrM supplementation did not affect mean 

slowest 10% RT (p = 0.694, Figure 3.3A), mean fastest 10% RT (p = 0.693; Figure 3.3B), number 

of lapses (p = 0.533; Figure 3.3C) or the mean RRT (p = 0.390; Figure 3.3D) following SR. 



31 
 

Figure 3.3: Psychomotor vigilance task performance at baseline and after SR (3 hours TIB) 
following 7 days of CrM supplementation for: (A) mean slowest 10% response time; (B) mean 
fastest 10% response time; (C) number of lapses; and (D) mean reciprocal response time. 
Values are presented as mean ± standard deviation. Dotted lines represent each participant.  

Table 3.3: P-values of main effects and interactions for PVT variables. 

 Lapses Mean RRT Mean 10% 
fastest RT 

Mean 10% 
slowest RT 

Test 0.599 0.013 0.017 0.602 
Condition 0.061 0.461 0.181 0.942 
Test x condition 0.533 0.390 0.693 0.694 

 

3.4.5 Mood, subjective sleepiness, and workload 

SR significantly reduced vigour (p < 0.001; Figure 3.4A), and increased fatigue (p = 0.004; Figure 

3.4B), subjective sleepiness (p = 0.009; Figure 3.4C) and subjective workload (p = 0.017; Figure 



32 
 

3.4D). However, CrM supplementation did not affect vigour (p = 0.818; Figure 3.4A), fatigue 

(p = 0.591; Figure 3.4B), subjective sleepiness (p = 0.858; Figure 3.4C) or workload following 

SR (p = 0.229; Figure 3.4D). 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 3.4: Mood, subjective sleepiness, and workload at baseline and after SR (3 hours TIB) 
following 7 days of CrM supplementation for: (A) vigour; (B) fatigue; (C) subjective sleepiness; 
and (D) workload. Values are presented as mean ± standard deviation. Dotted lines represent 
each participant. 

 

W
o

rk
lo

ad
 



33 
 

Table 3.4: P-values of main effects and interactions for mood, subjective sleepiness, and 
workload. 

 
Vigour Fatigue 

Subjective 
sleepiness 

Workload  

Test <0.001 0.004 0.009 0.017 
Condition 0.849 0.300 0.931 0.153 
Test x condition 0.818 0.591 0.858 0.229 

 

3.5 Discussion 

This study investigated the effects of a 7-day CrM supplementation protocol compared with 

placebo on cognitive performance, mood, subjective sleepiness, and perceived workload 

following a single night of SR (3 hours TIB). SR impaired some cognitive performance variables 

(mean RRT, mean 10% fastest RT), mood (vigour and fatigue), subjective sleepiness and 

workload; whereas other cognitive performance variables (number of lapses, mean slowest 

10% RT) remained unaltered. In contrast to our hypothesis, CrM supplementation did not 

mitigate the effects of SR on cognitive performance, mood, sleepiness, and workload following 

SR compared with the placebo condition. These observations suggest that increasing brain 

PCr, by ingesting 20 gday-1 of CrM for 7 days, is insufficient to ameliorate potential metabolic 

derangements and cognitive deficits associated a single night of SR.  

3.7.1 Effect of sleep restriction 

The single night of SR in the present study, which limited sleep from 0300 to 0600, impaired 

multiple aspects of PVT performance. Mean RRT decreased and mean fastest 10% RT 

increased, indicating mean speed and fastest responses were slower. In the present study, SR 

was not sufficient to increase the number of lapses, which have been shown to increase 

following multiple nights of SR (Belenky et al., 2003) or during total SD (van Dongen et al., 

2003). Consistent with the effects on cognitive performance were increases in subjective 

fatigue, sleepiness, and perceived workload, and a reduction in vigour. These observations 

suggest that our protocol impaired multiple aspects of cognition. Cognitive performance 

impairments could have been related to disruptions in energy metabolism, for which the 

efficacy of creatine loading was being assessed. 



34 
 

3.5.1 Effect of creatine supplementation following sleep restriction 

Creatine supplementation did not appear to affect any aspects of cognition following SR. 

Considering SR is likely to reduce brain PCr stores (Plante et al., 2014), CrM supplementation 

was expected to elicit a favourable effect due to its ability to increase brain creatine (Turner 

et al., 2015). As brain creatine stores were not measured in the present study, it is impossible 

to confirm whether the SR protocol was sufficient to elicit a metabolic effect which creatine 

loading could mitigate. For example, the benefits of creatine on cognitive performance and 

mood were not observed until 18 to 24 hours of SD (McMorris et al., 2006), suggesting 3 hours 

TIB may have been insufficient to disrupt energy metabolism for which creatine 

supplementation could ameliorate. Future studies should therefore include measures of brain 

energy metabolism to explain cognitive outcomes. 

The cognitive task load may also not have elicited sufficient metabolic stress in addition to SR 

for which the benefits of creatine could manifest. In the present study, the 10-minute PVT is 

known to enhance time-on-task effects (Lim & Dinges, 2008). However, while participants 

reported increased workload following SR, the PVT may not have been adequately demanding. 

For example, in studies demonstrating the beneficial effects of creatine supplementation at 

rest, Watanabe et al. (2002) used a 2 x 15-minute mathematical calculation task, and van 

Cutsem et al. (2020) used a 90-minute Stroop test. Moreover, in resting (van Cutsem et al., 

2020; Watanabe et al., 2002) and SD (McMorris et al., 2007; McMorris et al., 2006) studies 

assessing the effect of creatine supplementation on cognitive performance, tasks favourably 

affected by creatine supplementation were typically more complex than the PVT (e.g. random 

number generation task). As such, the effort required during shorter and simpler tasks was 

likely insufficient for observing a favourable effect of creatine supplementation on cognition 

following sleep loss. 

3.5.2 Brain creatine stores 

Since brain creatine concentration was not measured in the present study, the efficacy of the 

creatine dosing protocol (i.e. 20 gday-1) increasing brain creatine and PCr stores cannot be 

confirmed. As such, it is possible that brain creatine and PCr stores were not sufficiently 

increased to alter cognition, despite using a validated loading protocol. For example, 

Wilkinson et al. (2006) used the same dosing protocol and found no increase in brain creatine 

in 18 young sportsmen (mean age 22.7 years old). Additionally, the effect of creatine 



35 
 

supplementation on total brain creatine appears to vary between individuals (Dechent et al., 

1999) and according to age, gender and body weight (Dechent et al., 1999; Solis et al., 2017). 

Increases in brain creatine, compared to muscle, appear less consistent, which may be due to 

the brain being less sensitive to dietary and supplemental creatine sources (Dolan et al., 2019; 

Roschel et al., 2021). It is therefore possible that the creatine supplementation has failed to 

increase brain creatine in the present study. Future studies should aim to measure brain 

energy dynamics either via measurement of substrate or alternative measures of activation 

(e.g. neurophysiological or haemodynamic markers).  

3.5.3 Study limitations 

In addition to not measuring brain energy metabolism and creatine stores, the present study 

may have failed to identify effects of creatine supplementation due to the following 

methodological reasons. Firstly, the small sample size (i.e. n = 7) due to challenging 

recruitment amidst COVID restriction, meant the study was probably underpowered to detect 

the subtle effects on cognition; however, the study was a within-participant design (i.e. 

crossover), meaning each participant received both supplements (i.e. CrM and PLA), 

increasing the possibility of detecting an effect. Other creatine loading and sleep loss studies 

(McMorris et al., 2007; McMorris et al., 2006) were typically parallel design which compare 

two different groups, thus highlighting a strength of the present investigation. Secondly, as 

participants slept at home during the entire protocol, sleep periods could not be controlled as 

in laboratory-based studies and therefore there were some minor deviations with the 

protocols. Although participants sleeping significantly longer than 3 hours during the SR night 

were excluded from data analyses, sleeping in uncontrolled conditions could have increased 

variability in sleep duration and quality. Nevertheless, all participants were closely monitored 

via wrist actigraphy and sleep duration and waking time prior to testing did not differ between 

supplemental conditions.  

3.5.4 Conclusion  

In conclusion, a single night of SR (3 hours TIB) impaired aspects of PVT performance, 

increased fatigue, subjective sleepiness and perceived workload, and reduced vigour. Whilst 

CrM supplementation has previously been associated with favourable effects on cognition in 

stressful situations, such as sleep loss, the present study indicated that a 7-day CrM 

supplementation protocol does not mitigate the cognitive effects of a single night of SR (3 



36 
 

hours TIB). However, the current observations are preliminary and further research using 

larger sample sizes and multiple nights of SR, which is common in various occupational 

environments, is required.



37 
 

4 Conclusion and recommendations 

4.1 Study summary 

This double-blinded controlled, cross-over trial was designed to examine the effect of a 7-day 

CrM loading protocol (20 gday-1 of CrM or placebo divided into 4 equal 5 g doses) on cognitive 

performance, mood, subjective sleepiness, and perceived workload following a single night of 

SR (3 hours TIB). To the author’s knowledge, this is the first study to investigate the effect of 

CrM supplementation on cognitive responses following SR. 

This study was conducted under a hybrid setting of home and laboratory environments. Whilst 

all testing were performed in the laboratory, participants slept at home, including on the SR 

night. To ensure participants adhered to the study’s requirements at home, rigorous 

monitoring strategies were employed, which included text messages from participants 

reporting CrM ingestion, wrist-worn actigraphs and sleep diaries to monitor sleep duration 

and timing, and image-assisted diet records to monitor dietary intake. Caffeine was also 

prohibited, and physical activity limited in the 24 h prior to testing. Participants’ sleep 

duration, reported dietary intake and exercise load did not differ between the placebo and 

CrM conditions and were, therefore, not expected to have confounding effects on cognitive 

outcomes. As such, the study design was reasonably well controlled to elucidate if there was 

an effect of CrM.  

Several challenges were encountered during the study. This study started whilst the country 

was under COVID-related restrictions, making participant recruitment and data collection 

difficult. Unfortunately, two participants were infected with COVID during the intervention, 

meaning they were not able to test after the SR period and consequently their data could not 

be used. Additionally, two participants were excluded because they exceeded the sleep 

duration requirements on the SR night (i.e 3.71 and 5.53 hours). Due to a lack of time and 

availability, and unwillingness to repeat the study protocol, participants did not undertake the 

intervention and SR again; therefore, their data were also unable to be used. Due to the 

resulting low sample size (i.e. n = 7), the study was treated as a pilot study and these data 

considered preliminary.  



38 
 

The objectives of this research thesis were to examine whether 7 days of ingesting 20 gday-1 

of CrM affects aspects of cognition following a single night of SR (3 hours TIB), including: 

- Simple cognitive performance (vigilance and sustained attention), using a 10-minute 

version of the PVT (PVT-192). 

- Mood (fatigue and vigour), using a the validated computerised ANAM® mood scale. 

- Subjective sleepiness, using the Stanford Sleepiness Scale. 

- Perceived workload, using the NASA Task Load Index (NASA-TLX) 

 

4.2 Main findings 

4.2.1 Hypothesis 1: A single sleep restriction night (3 hours TIB) will negatively affect 

cognitive performance, mood, subjective sleepiness, and perceived workload. 

Cognitive performance, as measured by a 10-minute PVT, demonstrated some variables were 

degraded following SR (mean RRT and mean fastest 10% RT), whereas others were stable (i.e. 

number of lapses and mean slowest 10% RT). Further, SR impaired mood (i.e. increased fatigue 

and reduced vigour), and increased subjective sleepiness and perceived workload. These 

findings support the hypothesis by indicating that a single night of SR (3 hours TIB) can be 

sufficient to impair some aspect of cognition. 

4.2.2 Hypothesis 2: CrM loading will mitigate the adverse effect of a single night of 

sleep restriction night (3 hours TIB) on cognitive performance, mood, 

subjective sleepiness, and perceived workload. 

The absence of a test (i.e. baseline and SR) and supplementation (i.e. creatine vs placebo) 

interaction indicates that CrM loading did not moderate SR-related deficits in cognitive 

performance variables (i.e. lapses, mean RRT, mean slowest 10% RT and mean fastest 10% 

RT), mood (fatigue and vigour), subjective sleepiness, and perceived workload following SR (3 

hours TIB). However, further investigation is warranted as a small sample size and 

methodological limitations prevent making definitive conclusions.  

4.3 Study strengths 

1. This double-blinded study was conducted under a hybrid setting of home and 

laboratory environments. Whilst participants slept at home, participants were 

monitored for key variables that may influence the study’s primary outcomes, such as 



39 
 

sleep duration and diet, and requirements standardising physical activity were 

implemented. The testing environment was strictly controlled, and participants were 

provided with a quiet room with no distraction and constant temperature. 

Participants’ testing time between baseline and following SR for both interventions 

was kept the same, thus eliminating potential time of day effects.  

2. A 10-minute PVT was used, which is a valid and reliable task to measure reaction time 

and vigilance following sleep loss. The PVT is widely used for measuring these variables 

in sleep studies as it is sensitive to the effects of sleep loss. The PVT has been used in 

controlled laboratory and fields setting. Also, the PVT reduces intra-subject variability 

as it has minimal practice effects (Lim & Dinges, 2008, 2010). Participants were 

familiarised in the laboratory with the 10-minute PVT test prior to both interventions 

to minimise any possible practice effects (Balkin et al., 2004).  

3. To reduce the likelihood of abnormal sleep habits influencing outcome measures, 

participants were selected based on their habitual sleep pattern (i.e. typically sleeping 

from 2100 and 0000 to 0700 and 0900), sleep quality measured using the Pittsburg 

Sleep Quality Index score, and chronotype measured using the Horne-Östberg 

Morningness/Eveningness Questionnaire. If a participant had a Pittsburg Sleep Quality 

Index score >5 or scored to be definitely a morning or an evening chronotype, they 

were not eligible for the study. Participants’ sleep duration at baseline (2 days), during 

the 7-day intervention and during the SR night were monitored by wrist actigraphy and 

sleep diary to ensure these sleeping patterns were maintained and participants 

adhered to the SR requirements. Average sleep duration did not differ between 

placebo and CrM interventions, thereby any confounding effects on cognitive 

performance and subjective variables were not expected following SR. 

4. The participant inclusion and exclusion criteria were established to minimise potential 

confounding effects on cognitive responses. Participants were recruited with strict 

inclusion and exclusion criteria to ensure participants were healthy, within an age 

range of 18-50 years, had a BMI within the healthy weight range (i.e. <30 kgm²), were 

free from chronic illness and sleep disorders, consumed <3 cups caffeinated beverages 

per day, and were not involved in shift work or trans-meridian travel within the 28 days 

prior to the study to reduce possible confounding effects on cognition. 



40 
 

5. Participants and researchers communicated daily during each intervention to ensure 

participants complied with sleep, supplement, and dietary requirements. 

6. Linear mixed effect models were used to analyse data, which included a random 

intercept for participant and a random slope for test. These are more appropriate than 

basic linear models due to: 1) the repeated-measures crossover design, which means 

data are non-independent; 2) the potential for individual variability; and 3) the effect 

of SR could vary within each individual between supplement conditions. Whereas basic 

linear models assume each data point is discrete and does not account for variability 

in individual responses and the effect of sleep restriction.  

4.4 Study limitations 

1. The initial aim was to recruit 15 to 20 participants; however, due to the difficulty of 

recruiting participants (e.g. COVID-related restrictions, and strict eligibility/exclusion 

criteria) and adhering to the challenging protocol, and the limited time frame to collect 

data, only a small sample size of n = 7 could be attained. This reduced the power of the 

study to detect a statistically significant effect of CrM on cognition following SR and, 

therefore, the data in considered preliminary and the study was reduced to a pilot 

study. 

2. Participant characteristics in this study do not reflect all demographics within the New 

Zealand adult population. Therefore, the results cannot be directly extrapolated to 

other populations, such as shift workers, older adults, overweight-obese individuals, 

anxious, depressed, and stressed individuals, smokers, high caffeine consumers and 

individuals with abnormal sleep habits.  

3. The efficacy of the CrM loading protocol on increasing brain creatine content could not 

be confirmed as pre- and post- brain creatine content was not measured. It is possible 

that the dosing protocol failed to increase brain creatine stores, which could explain 

why no effect was found after CrM loading. Honest reporting from participants to take 

the supplements was expected; however, it is not ascertained that participants took 

the supplements as instructed (i.e. 5 g 4 times per day for 7 days).  

4. Sleep and SR took place at home, meaning participants were not as closely monitored 

as what occurs in laboratory-based studies. However, participants were monitored 

with actigraph watches and those not compliant with the restrictions were detected. 



41 
 

Further, as participants were sleeping in a familiar environment, it is more likely they 

adhered to their normal sleep patterns and ascertain higher quality sleep. Whereas, 

sleeping in uncontrolled conditions would have increased the potential for poor sleep.  

5. The PVT is a valid and reliable measure of cognitive performance during SR (Lim & 

Dinges, 2008, 2010); yet, the 10-minute version of the PVT may have not been 

demanding enough to elicit sufficient metabolic stress in addition to SR for which the 

benefits of creatine could manifest. 

6. Physical activity was limited, and caffeine consumption forbidden before testing to 

limit the known effects to mask sleep loss related deficits (Irwin et al., 2020; McLellan 

et al., 2016; Temple et al., 2018; Tomporowski, 2003). However, adherence was not 

monitored, and we relied on participants’ honesty. 

4.5 Final recommendations 

4.5.1 Impact 

The study findings highlight that a single night of SR can negatively affect cognition, mood, 

sleepiness, and perceived workload. Sleep loss increases unintentional human error and can 

lead to dramatic outcomes. Given the high prevalence of SR in our society, including in the 

military, medical care, transport, and search and rescue, strategies to mitigate the effects of 

sleep loss are critical. Previous research has demonstrated the beneficial effects of CrM 

supplementation on cognitive performance when brain metabolism is impaired, such as 

following sleep loss; however, little is known about the effect of CrM on cognition, mood, 

sleepiness, and perceived workload following a single night of SR. Although this study was 

unable to demonstrate any effect of CrM supplementation on cognitive responses following 

SR, it can provide methodological considerations and recommendations for further research. 

4.5.2 Recommendations 

The f