Putting the focus on sleep quality: subjective sleep measures, actigraphy and brain correlates in ageing

Universidade do Minho Escola de Medicina Liliana Patrícia Carvalho Amorim setembro de 2019 Putting the focus on sleep quality: subjective sleep measures, actigraphy and brain correlates in ageing Liliana Patrícia Carvalho Amorim Putting the focus on sleep quality: subjective sleep measures, actigraphy and brain correlates in ageing UMinho|2019 Liliana Patrícia Carvalho Amorim setembro de 2019 Putting the focus on sleep quality: subjective sleep measures, actigraphy and brain correlates in ageing Trabalho efetuado sob a orientação da Doutora Nadine Correia Santos e do Professor Doutor Nuno Jorge Carvalho Sousa Tese de Doutoramento Doutoramento em Ciências da Saúde Universidade do Minho Escola de Medicina ii DIREITOS DE AUTOR E CONDIÇÕES DE UTILIZAÇÃO DO TRABALHO POR TERCEIROS Este é um trabalho académico que pode ser utilizado por terceiros desde que respeitadas as regras e boas práticas internacionalmente aceites, no que concerne aos direitos de autor e direitos conexos. Assim, o presente trabalho pode ser utilizado nos termos previstos na licença abaixo indicada. Caso o utilizador necessite de permissão para poder fazer um uso do trabalho em condições não previstas no licenciamento indicado, deverá contactar o autor, através do RepositóriUM da Universidade do Minho. Licença concedida aos utilizadores deste trabalho iii Acknowledgments I would first like to express my sincere gratitude to my supervisors, Nadine Santos and Professor Nuno Sousa, for challenging me and providing me the opportunity to learn a great amount of skills beyond my area of expertise. I also want to acknowledge the ICVS colleagues, who contributed to my growth during this PhD journey, and that took time to share ideas and discuss science with me. A special thanks to Miguel Pais-Vieira, Hugo Almeida, João Sousa, Patrício Costa and Mónica Gonçalves. I also want to acknowledge a very special group of people that made my PhD “sunnier”. Thank you: Teresa and Belina, my anchors; Sofia, my “social manager” and companion in the cortisol adventures; Madalena, my ”Wednesdays’ of sushi” partner; Paulinho, Pedro, Ricardo and Carlos for all the pacience, help and partnership; Rui, Luís, Patrícia e Rita for all the laughs and conversations; Margarida, Eduardo, Catarina, Dinis, Sandro, Gabriela, Ricardo and Sónia, for the friendship, the brunchs and afternoons of games. Susana, Patrícia, São, Cassandra and Ângela, thank you for the “rainbows” in rainy days. I also want to acknowledge my family; thank you for being able to take the absences with patience and provide me with all the love and understanding that I needed! Thank you João, for the love, late night discussions and for pushing me out of my confort zone; this journey would not have been the same without you! The work presented in this thesis was performed in the Life and Health Sciences Research Institute (ICVS), Minho University and the Clinical Academic Center (2CA), Hospital of Braga. Financial support was provided by grants from the Foundation for Science and Technology (FCT) through the PhD grant SFRH/BD/101398/2014, by FEDER funds through the Operational Programme Competitiveness Factors - COMPETE and National Funds through FCT - Foundation for Science and Technology under the project POCI-01-0145-FEDER-007038; and by the projects NORTE-01-0145-FEDER-000013 and NORTE-01-0145FEDER-000023, supported by Norte Portugal Regional Operational Programme (NORTE 2020), under the PORTUGAL 2020 Partnership Agreement, through the European Regional Development Fund (ERDF). Financial Support was also obtained from “SwitchBox” (Contract HEALTH-F2-2010-259772), co-financed by the Portuguese North Regional Operational Program (ON.2 – O Novo Norte), under the National Strategic Reference Framework (QREN), through the European Regional Development Fund (FEDER), and by the Fundação Calouste Gulbenkian (Portugal) (Contract grant number: P-139977; project “Better mental health during ageing based on temporal prediction of individual brain ageing trajectories (TEMPO)”). “To go far you must begin near, and the nearest step is the most important one.” (Jiddu Krishnamurti) iv STATEMENT OF INTEGRITY I hereby declare having conducted this academic work with integrity. I confirm that I have not used plagiarism or any form of undue use of information or falsification of results along the process leading to its elaboration. I further declare that I have fully acknowledged the Code of Ethical Conduct of the University of Minho. v Putting the focus on sleep quality: subjective sleep measures, actigraphy and brain correlates in ageing Abstract Sleep is a multidimensional phenomenon with a relevant role in the maintenance of organism homeostasis, overall well-being and optimal cognitive function. Throughout the lifespan, changes in sleep dimensions occur, namely on its timing, duration, architecture and quality. In what concerns to sleep quality, studies are highly heterogeneous; not only the tools used are extremely variable, but also the way this concept is defined. Thus, the present thesis aims at providing a thorough analysis on the meaning of sleep quality, contributing to the clarification of its definition. It also aims to determine the sleep patterns, routines and quality of Portuguese community-dwellers across the adult lifespan and determine the main predictors of self-reported sleep. It is also an aim to explore the association between a composite measure of sleep quality and brain correlates. For these purposes, a systematic review of the literature on sleep quality meaning was performed and three original studies developed considering a cross-sectional and/or a longitudinal approach. Portuguese community-dwellers within the adult lifespan (18 and more years) were recruited and self-reported sleep quality and psychological variables were assessed in the first two studies and in the third, neuroimaging information was also collected. Results show that sleep quality is a multidimensional concept that should integrate information from different settings (e.g. clinical measures such as the Pittsburgh Sleep Quality Index, and parameters that reflect the interpretation of lay individuals of it). Lay people interpretation of sleep quality seems to be stable across the adult lifespan and no differences between men and women are observed regarding the parameters reported. Poor subjective sleep quality is associated with decreases in functional and structural connectivity of specific brain networks, with an overlapping node in the middle left temporal region. Overall, the present work contributes for the clarification of sleep quality concept, enabling a better comparison between study results. Despite the contributions of this doctoral work to the body of literature, there are still avenues to be explored so that individuals can benefit from personalized sleep care, and physicians can better act on the subjective informations reported by their patients, feasibly taking advantages of the new technologies to better monitor sleep in ecological settings. Key Words: Actigraphy; MRI; PSQI; Psychological Variables; Subjective sleep quality. vi Colocando o foco na qualidade de sono: medidas subjetivas de sono, actigrafia e correlatos cerebrais no envelhecimento Resumo O sono é um fenómeno multidimensional, com um papel relevante na manutenção da homeostasia do organismo, no seu bem-estar geral e no bom funcionamento da sua função cognitiva. Ao longo da vida ocorrem alterações nas suas diferentes dimensões, nomeadamente nos seus horários, duração, arquitetura e qualidade. No que respeita à qualidade de sono, os estudos são altamente heterogéneos; não só os instrumentos utilizados são muito variáveis, mas também a forma como o conceito é definido. Assim, a presente tese tem como objetivos realizar uma análise aprofundada do conceito “qualidade de sono” e contribuir para a clarificação da sua definição; determinar os padrões, rotinas e a qualidade de sono de uma amostra de indivíduos potugueses adultos, bem como os principais preditores das variáveis de sono reportadas; e determinar a associação entre uma medida compósita de qualidade de sono e correlatos cerebrais. Para este propósito, realizaram-se uma revisão sistemática da literatura sobre a definição de qualidade de sono e três estudos empíricos com uma abordagem transversal e/ou longitudinal, onde se recrutaram indivíduos de 18 e mais anos, e se avaliaram a qualidade de sono e algumas variáveis psicológicas nos dois primeiros estudos, e se realizou ainda uma ressonância magnética cerebral no terceiro. Os resultados indicam que a qualidade de sono é um conceito multidimensional, que deve integrar informação de de diferentes contextos (ex. medidas clínicas como a do Índice de Qualidade de Sono de Pittsburgh, mas também parâmetros que decorrem da interpretação do conceito pelo indivíduo). A interpretação deste conceito pelos indivíduos da comunidade parece ser estável ao longo da vida adulta, não se observando diferenças entre homens e mulheres nos parâmetros reportados. Uma má qualidade de sono correlaciona-se com uma diminuição da conectividade funcional e estrutural em redes cerebrais que se sobrepõem na região temporal média esquerda. O presente trabalho contribui para a clarificação da definição de qualidade de sono, promovendo comparações mais adequadas entre resultados de estudos. No entanto, apesar das contribuições, existem ainda caminhos a serem explorados, para que os indivíduos possam beneficiar de cuidados de sono personalizados e para que os seus médicos possam agir melhor sobre as informações subjetivas relatadas, aproveitando as vantagens das novas tecnologias para melhor monitorizar a qualidade de sono em ambientes ecológicos. Palavras-chave: Actigrafia; MRI; PSQI; Qualidade Subjetiva de Sono; Variáveis psicológicas. vii TABLE OF CONTENTS Acknowledgments iii Financial support iii Abstract v Resumo vi TABLE OF CONTENTS vii List of Abbreviations x List of Figures xii List of Tables xiv CHAPTER I 16 STATE OF THE ART 1.1. Introduction 17 1.2. Historical perspective 18 1.3. Sleep architecture 21 1.5. A Sleep Regulation Model 25 1.6. Sleep and neuroimaging 28 1.7. Sleep and ageing 30 1.8. Epidemiology of Sleep 32 1.9. Sleep quality and its measurement 34 Objectives 39 References 41 CHAPTER II 51 THE IMPORTANCE OF DEFINING WHAT WE ARE MEASURING: A SYSTEMATIC REVIEW ON SLEEP QUALITY MEANING Abstract 53 Introduction 54 Methods 55 Results 56 viii Discussion 85 Conclusions 89 Conflict of interest statement 89 Acknowledgments 89 References 90 CHAPTER III 102 SLEEP QUALITY MEANING AND ITS ASSOCIATION WITH THE PITTSBURGH SLEEP QUALITY INDEX (PSQI): A POPULATION-BASED STUDY ACROSS THE ADULT LIFESPAN Abstract 104 Introduction 105 Methods 107 Results 112 Discussion 125 Conclusion 130 Conflict of interest statement 131 Contributions 131 Acknowledgments 131 References 132 Appendix 1 135 CHAPTER IV 136 A SHORT REPORT ON WEEK-WEEKEND VARIABILITY OF SELF-REPORTED SLEEP PATTERNS, ROUTINES AND QUALITY: PRELIMINARY RESULTS Abstract 138 Introduction 139 Methods 139 Results 141 Discussion 146 Conflict of interest statement 147 Acknowledgments 147 xv Table 3. Longitudinal characterization of sleep patterns characterization of the subsampled cohort (n=53). CHAPTER V Table 1. Cohort characterization in terms of socio-demographic factors and clinical and psychological parameters. Table 2. Correlation between subjective sleep quality and demographic, clinical and psychological parameters. Table 3. Differences regarding age, PSQI global score, GDS and ESS between the original cohort and the subsample that used actigraphy. Table 4. Cohort characterization in terms of actigraphy variables. Table 5. Correlation between subjective sleep quality, sleepiness, depressive symptoms and actigraphic parameters. Table 6. Average integrated graph properties. Table 7. Functional and Structural Connectivity results for the tested thresholds. Table 8. Regional brain volumes association with age, sex, subjective sleep quality and depressive symptoms. 16 CHAPTER I The state of the art 17 1.1. Introduction Sleep is a ubiquitous phenomenon, critical for the organism homeostasis and overall well-being. Across the lifespan, alterations in sleep habits, routines and architecture occur and often, there is an increase in sleep complaints, especially in middle-aged and older individuals. Notwithstanding, it is still unclear whether these changes are biologically programed or a consequence of external conditions (e.g. societal burdens or/and chronic sleep restriction throughout the years). Indeed, while some studies have shown that it is not the ageing process per se that deteriorates (at least some) sleep processes (e.g. the ability to fall asleep), it is still to be determined how many sleep complaints and of what type should be expected in normative ageing. Thus, and because there is an important behavioral-related component associated to sleep, it is crucial to determine how individuals integrate and provide meaning to these sleep changes. In fact, determining the specificities of sleep complaints can be a relevant contribute to better understand and overcome some sleep disturbances and to promote overall health, quality of life and the maintenance of a proper neuropsychological status. Thus, the aims of the present thesis are twofold. First, to clarify the subjective sleep quality concept by providing a critical view on this construct. Second, to present an integrative view on sleep patterns and subjective quality across the lifespan, exploring for associations between subjective sleep quality, psychological factors and brain correlates. This information is provided across five chapters. In the present chapter, Chapter I, a frame of the topic of the thesis provides the necessary theoretical background. Chapter II concerns a systematic literature review on sleep quality construct, focusing on subjective sleep quality and addressing needed clarifications on its meaning. Chapter III delivers a characterization of the rest patterns of community-dwellers from 18 to 87 years of age and addresses some of the limitations identified in the literature review presented in Chapter II. Associations between what is reported by the individual in terms of sleep quality, and the results from a standard self-reported measure of sleep quality (obtained from the Pittsburgh Sleep Quality Index), are also determined. In Chapter IV, differences between week and weekend days regarding sleep patterns, routines and quality are addressed, as well as age and sex differences in these patterns. Chapter V explores the association between the composite measure of sleep quality derived from the Pittsburgh Sleep Quality Index and structural and functional brain correlates. Chapter VI provides a summary of the findings of the thesis and discusses ongoing work and future prospects. 18 1.2. Historical perspective 1 Sleep is a recurrent and reversible neurobehavioral state characterized by behavioral quiescence, perceptual disengagement from the environment and decreased consciousness (Carskadon & Dement, 2011; Siegel, 2009). Unlike other behaviors which are directed towards a clear goal (e.g. getting food, mating, catching or running from something), sleep does not readily indicate any ‘intent’ of engaging in a clear action. Furthermore, what seems to remain upon a sleep episode is a subjective experience of loss of consciousness and not a proper remembrance of the occurrence. Consequently, different individuals, throughout time and societies, can adopt different patterns and interpretations of sleep. In fact, for a long time, the lack of memory of mental activity during the sleep episode seemed to foster the notion of sleep as a state of reduced or non-existent brain activity (Scott & Sherrington, 2009) and, thus, not compatible with behaviors that nourish and/or propagate species. This conception was challenged with the discoveries of the regular cyclic alteration of rapid eye movement (REM) and non-REM (NREM) sleep (Aserinsky & Kleitman, 1953), and the strong association between REM sleep and vivid hallucinatory dreaming (Dement & Kleitman, 1957). Indeed, these served as evidence against this belief that sleep was caused by/or associated with a cessation of brain activity (Hobson, 2005). Beyond its biological component, sleep is also a social, cultural and historically variable phenomenon (Ohayon, 2004; Williams, Meadows, & Arber, 2010). Hence, changes in each of these domains can relate to changes in sleep patterns. A prime example of this occurred during the Industrial Age, when sleep went from being regarded as ‘the honey-heavy dew of slumber’ (Shakespeare), or ‘the golden chain that ties health and our bodies together’ (Thomas Dekker), to the ‘criminal waste of time and a heritage from our cave days’ (Thomas Edison). In fact, while before the Industrial Age it was customary to receive visitors in bed, thereafter the bedroom was regarded as private and sleeping for seven or eight hours per night as laziness1. Interestingly, even nowadays, after the importance of sleep for health and well-being has been demonstrated, there are still certain sayings, such as ‘don’t get caught napping’, ‘if you snooze you lose’ or ‘time is money’, that promote this view of sleep as either optional, a luxury, or unimportant (Colten et al., 2006). More so, the advent and widespread use of artificial light also changed human behavior. 1 Most information from this section was obtained from the site http://healthysleep.med.harvard.edu/interactive/timeline 19 Not only work and leisure hours were extended past natural sunset, but also a significant segment of the daytime workforce was moved indoors, limiting the exposure to natural light (Sharkey & Van Reen, 2014). This extension of work and leisure hours past natural sunset has been shown to associate with poorer mental health and more sleep disturbances in individuals working longer hours (Afonso, Fonseca, & Pires, 2017; Virtanen et al., 2009). Similarly, “day-to-day” social factors also seem to affect sleep across the lifespan, including marriage/co-habitation (Troxel et al., 2010), parenthood (Medina et al., 2009; Smedje et al., 1998), caregiving (Castro et al., 2009; McCurry et al., 2007; Rittman et al., 2009; Simpson & Carter, 2010; Song et al., 2018; von Känel et al., 2012) or widowhood (Monk et al., 2008). Thus, sleep disruption appears to be on the rise in our 24-h society, albeit its important for a large set of biological and psychological processes (Cirelli & Tononi, 2008; Diekelmann & Born, 2010; Irwin, 2015; Irwin & Opp, 2017; Maquet, 1995; Meerlo et al., 2009; Payne et al., 2009; Pilcher, Ginter, & Sadowsky, 1997; Shokri-Kojori et al., 2018; Stickgold, 2006; Tononi & Cirelli, 2006; Walker & Stickgold, 2004; Wulff et al., 2010; Xie et al., 2013). Therefore, considering the different features of sleep, namely duration, timing and quality, understanding and determining the conceptions and expectations of the individual towards sleep is crucial to demystify myths and choose the most effective approach to promote a good sleep and of quality. In Figure 1, a chronological frieze of some of the most relevant milestones regarding sleep and its research is presented. 20 Figure 1. Chronological frieze of sleep research milestones across time (based and adapted from: http://healthysleep.med.harvard.edu/interactive/timeline). 21 1.3. Sleep architecture To access the basic structural organization of sleep, the standard method used is Polysomnography (PSG), which allows for sleep characterization in its discrete stages through a minimum of eleven channels, including electroencephalogram (EEG), electromyogram (EMG), electrooculogram (EOG), oxygen saturation (SpO2), and electrocardiogram (ECG). The quantitative electroencephalogram (EEG) mapping has shown that during the different sleep phases there are brain regional differences in electrical activity (Achermann & Borbély, 1997, 1998) and sleep has been classified as either rapid-eye-movement (REM) or non-REM sleep based on these changes in EEG, EOG and EMG (Figure 2). Figure 2. Features of wake-sleep behavior states (adapted from Hobson, 2005 with permission; license number: 4553100288346). In 1968, Rechtschaffen and Kales provided the first frame for sleep scoring. Although these guidelines were meant just as a reference, they rapidly became gold standard (Himanen & Hasan, 2000). The reappraisal of the scoring occurred only in 2007, with not only the alteration of the standard guidelines for sleep classification, but also the development of guidelines for terminology, recording method, and scoring rules for sleep-related phenomena by the American Academy of Sleep Medicine (AASM) (Iber, Ancoli-Israel, Chesson, & Quan, 2007). Currently, sleep is divided into wake, NREM stage 1 (N1), NREM stage 2 (N2), NREM stage 3 (also denominated ’deep sleep’, N3 reflects slow wave sleep and stages S3 + S4 in the previous scoring from 1968) and REM sleep (Figure 3). NREM sleep is characterized by an electrical activity progressively more synchronous 22 and with slower waves of bigger amplitude; while REM sleep is characterized by rapid eye movements, asynchronous cortical electric activity with low amplitude waves of high frequency, muscular hypotony (except for diaphragmatic muscles and muscles from middle ear and erectile tissues) and dreamlike activity (Paiva & Penzel, 2011; Stickgold & Walker, 2010). Furthermore, within NREM sleep, N1 is considered a transition period from wake to sleep, N2 is characterized by typical sleep elements (such as sleep spindles and K-complexes), and N3 is described as a deep sleep with slow electric waves of high amplitude (Paiva & Penzel, 2011) (Figure 3). REM sleep, due to its low-amplitude, high mixed-frequency waves, has been termed ‘desynchronized’ sleep (Stickgold & Walker, 2010). Figure 3. EEG trace for the different stages of sleep. Overall, it is expected that an adult without any sleep disturbance should be able to have a night of about 8 hours of sleep (Hirshkowitz et al., 2015), which can represent 4 or 5 sleep cycles, each of 90 to 120 minutes (Stickgold & Walker, 2010). The way REM, NREM and wake are distributed within each of these cycles, varies with the part of the night in which it occurs and with the previous sleep episode (Dement & Kleitman, 1957; Stickgold & Walker, 2010). For example, in an individual sleeping within the appropriate quality and range of time, NREM sleep has a longer duration in the first half of the night, while, in the second half of the night, REM sleep is predominant (Figure 4). Wake NREM Sleep – Stage N1 NREM Sleep – Stage N2 1 second Sleep Spindles Complex K NREM Sleep – Stage N3 REM Sleep 23 Figure 4. Hypnogram depicting the different sleep stages of a healthy individual, over 8-h nocturnal sleep (image from Blume, del Giudice, Wislowska, Lechinger, & Schabus, 2015). 1.4. Neurobiology of sleep As already mentioned, the behavioral hallmarks of sleep are diverse (ranging from reduced responsiveness to external stimuli and rapid reversibility, to homeostatic rebound after sleep loss). While early animal studies shed light on the importance of the reticular brainstem region for sustained wakefulness (Magoun, 1952), later studies have shown that the majority of sleepregulating stimuli are integrated in the ascending reticular activating system (ARAS) originated in the brainstem (Saper, Scammell, & Lu, 2005). Despite the exact wake-sleep mechanisms are still not completely understood or defined, it is consensual that a sleep control system has to be able to fulfill a multitude of functions, among which blocking locomotor activity, gating sensory pathways, inhibiting arousal systems and relieving sleep pressure (Donlea et al., 2018). The first hypothesis on wake-sleep mechanisms and its neuro-circuitry appeared around 1930’s. Based on the observation of patients with the viral illness encephalitis lethargica, Baron Constantin von Economo proposed a wake-promoting influence from the brainstem that would keep the forebrain awake, and a sleep-promoting influence from the anterior hypothalamus that would contradict the waking drive during sleep. In the later decades of the 20th century, it was found that the wake-promoting influence innervated the thalamus and passed through the more ventrally situated hypothalamus and the basal forebrain, to project to the entire cerebral cortex (Saper et al., 2005). Specifically, neurons from monoaminergic cell groups project to the intralaminar and midline thalamic nuclei and innervate the lateral hypothalamus, basal forebrain, and cerebral 24 cortex (Saper et al., 2005). These neurons fire faster during wakefulness compared to non-REM (NREM) sleep, and most stop firing altogether during rapid eye movement (REM) sleep (AstonJones and Bloom, 1981; Fornal et al., 1985; Steininger et al., 1999). At this time, it was considered that the wake-promoting influence arised mainly from monoaminergic and cholinergic neurons in the upper brainstem (Saper et al., 2005). However, later on, a sleep-promoting cell group was identified in the ventrolateral preoptic (VLPO) and median preoptic nuclei (Saper et al., 2010; Suntsova et al., 2002). This group of cells were shown to provide GABAergic innervation of the entire wake-promoting system, thus allowing the inhibition of arousal during sleep (Saper et al., 2010; Suntsova et al., 2002). Furthermore, evidence from lesion and optoand chemogenetic excitation and inhibition studies showed that monoaminergic and cholinergic systems played mainly a modulatory role on wake–sleep control. In fact, the backbone of the wake–sleep regulatory system seems to depend upon fast neurotransmitters, such as glutamate and GABA (gammaaminobutyric acid) (Saper et al., 2010; Saper et al., 2005) (Figure 5). Figure 5. Wake-sleep regulating systems. (adapted from Saper & Fuller, 2017 with permission. License number: 4540261429584). (a) In red, the fast neurotransmitter system that appear to play the largest role in promoting wakefulness; in brown, the monoaminergic, cholinergic, and peptidergic neurons in the brainstem and hypothalamus that have a modulatory role; in violet, two populations of GABAergic neurons in the lateral hypothalamus (LH). (b) In purple, the fast neurotransmitter systems that contribute to sleep promotion. The ventrolateral preoptic (VLPO) and median preoptic (MnPO) GABAergic neurons send axons to most components of the arousal system (in red, orange, and green), and are thought to inhibit them in a coordinated fashion. Abbreviations: 5HT, serotonin; Ach, acetylcholine; Hist, histamine; LC, locus coeruleus; LDT, laterodorsal tegmental nucleus NA, noradrenlaine; ORX, orexin; TMN, tuberomammillary nucleus. a b 31 population, it should be clear for the practicing physician, what changes to expect in sleep–wake patterns in normative ageing. On this, studies have reported that advancing into the fifth decade of life can lead to: (1) earlier bedtimes and rise times (i.e. advanced sleep timing); (2) longer time taken to fall asleep (i.e. longer sleep-onset latency); (3) shorter overall sleep duration; (4) increased sleep fragmentation (i.e. less consolidated sleep with more awakenings, arousals, or transitions to lighter sleep stages); (5) more fragile sleep (i.e., higher likelihood of being woken by external sensory stimuli); (6) reduced amount of deeper NREM sleep known as slow wave sleep (SWS); (7) increased time spent in lighter NREM stages 1 and 2; (8) shorter and fewer NREM-REM sleep cycles; and (9) increased time spent awake throughout the night (Mander et al., 2017) (Figure 8). An increase in the frequency of diurnal naps, particularly unplanned naps, is also observed in later life, given excessive sleepiness (Foley et al., 2007). However, as Mander and colleagues reported, excessive daytime sleepiness and daytime napping are not a universal feature of old(er) age. For a portion of older adults, daytime sleep propensity and daytime ratings of subjective sleepiness diminish with the transition from midlife into older adulthood (Mander et al., 2017; Dijk et al., 2010). On this, a factor that appears to be determinant is the presence of comorbid conditions, such as chronic pain, depression and sleep disorders, or frequent nighttime urination breaks (Mander et al., 2017; Foley et al., 2007; Vitiello, 2009). These changes can be interpreted in light of the reported functional deterioration of different systems, among which is the one involving the circadian clock. Thus, early intervention is critical and sleep-wake patterns might have an important role as a marker of disease onset or progression. For example, considering Alzheimer’s disease (AD), and the work from Lucey and colleagues (2019), where sleep patterns were analyzed in normal cognitive ageing (subjects without cognitive-related pathology), results showed that NREM sleep negatively correlated with tau pathology and Aβ deposition in several brain areas. Such suggests that alterations in NREM sleep may be an early indicator of AD pathology, and that noninvasive sleep analysis might be useful for monitoring patients at risk for developing AD (Lucey et al., 2019). Overall, there are several reasons for the difficulty of studying sleep in ageing. For instance, the high prevalence of comorbidities in older individuals raises questions about what should be considered within normal ageing. In fact, if on the one hand, individuals with comorbidities should be excluded, on the other, how representative or generalizable would be a sample composed only by “super-healthy” individuals? This debate is still ongoing and what should be considered “healthy 32 ageing” is still an open question. Nonetheless, studies of community-dwellers, within the normative ageing process, are needed in order to better understand sleep process and its mechanisms. This is even more important when considering some sleep research protocols that not only are timeconsuming, but also make it difficult to assess vulnerable populations. Furthermore, clinical samples can reflect referral biases that are complicated to overcome in order for the results to be generalized to the overall population. Thus, there is a posing need to start earlier the study of sleep in ageing in order to understand the different profiles and trajectories. 1.8. Epidemiology of Sleep In 1998, a representative cohort of Portuguese individuals with 18 and more years was interviewed by telephone about their sleep habits and disturbances (Ohayon, 2004; Ohayon & Paiva, 2005). Results showed that while approximately 12% of the sample (more females than males) reported difficulties in initiating sleep, 21% had difficulties in maintaining their sleep (again, more females compared to males). This trend increased with age; but, interestingly, for difficulty in initiating sleep, age evolution was in a U shape - higher for younger individuals (18–24 years), lower between the ages of 25 and 44, increasing again for the other age groups (Ohayon & Paiva, 2005). Results also showed that 9.8% indicated having non-restorative sleep, 28.1% reported having at least 1 insomnia symptom and 10.1% being globally dissatisfied with their sleep (Ohayon & Paiva, 2005). In what concerns sleep patterns, analysis were performed only for individuals with ages  55 years (Ohayon, 2004). Data indicated that the median sleep duration for individuals between 55 and 74 years old was 7h, and this value increased with age (7,5h for individuals between 75 and 84 years, and 8h for individuals with 85 or more years) (Ohayon, 2004). It was also observed that sleep latency remained the same across age groups (median=15 minutes) and that bedtime decreased with age (Ohayon, 2004). Specifically, individuals between 55 and 64 years went to bed in median at 23:30; while, individuals with ages comprised between 65 and 84 years, and with 85 years or more, went to bed in median at 23:00 and 22:45, respectively (Ohayon, 2004). In terms of waking time, the median wake time was 7:00 between 55 to 74 years, 7:30 in individuals with ages comprised between 75 and 84 years and 8:00 for individuals with 85 and more years (Ohayon, 2004). In 2006, Paixão and colleagues showed the evolution of sleep parameters in the adult Portuguese population (18 and more years of age) through a longitudinal study, with data collected in 1999 33 and 2004. In terms of amount of hours slept in weeknights, results showed that 87% of respondents in 1999 and 85% in 2004 usually slept six or more hours of nocturnal sleep in weekdays. Interestingly, the individuals reporting sleeping more hours during the week were most frequently men (88.4%, in 2004), from younger age groups and with higher educational levels. It was also observed that from 1999 to 2004, there was a decrease in the percentage of individuals sleeping six or more hours in Lisboa and Vale do Ave region but not in other regions. More so, this decrease trend was also observed with age. Results also show that most individuals did not have a nap routine (82% in 1999 and 86% 2004), and such habit was a more frequent behavior in older age groups. Furthermore, while in 1999, 45% of the individuals usually woke up during the night more than once a week, in 2004 that percentage raised to 71%. In terms of difficulties in falling asleep, percentages remained stable (19%) (Paixão, Branco, & Contreiras, 2006). Regarding the consequences of a poor sleep, 15% of the inquired in 1999 and 16% in 2004 complained of tiredness when waking up and 14% in 1999 and 12% in 2004 complained of diurnal sleepiness. In this line, approximately 11% of the participants in 1999 and 14% in 2004 reported the use of medication (Paixão et al., 2006), which follows the global trend of increase in sleep medication consumption (Bertisch, Herzig, Winkelman, & Buettner, 2014; Marom, Rennert, Stein, Landsman, & Pillar, 2016). Moreover, it was interesting to observe differences between the different country regions, with Algarve having the least percentage of individuals reporting the use of medication to sleep. A statistically significant increase was observed from 1999 to 2004 in the Northern region of Portugal, which went from being a region with a lower percentage of medication use to the highest. A statistical significant association was also found between waking up tired/diurnal somnolence all the time or most of the time and consumption of sleep medication (Paixão et al., 2006). Several factors can contribute for the disruption of the organism homeostatic balance and it seems that young adults, particularly college students, and older individuals, are particularly vulnerable groups to circadian rhythm/sleep disruptions. In fact, studies have shown that anywhere between 20 to 60% of college students are poor sleepers (Ahrberg et al., 2012; Bahammam et al., 2012; Lund et al., 2010; Preišegolavičiūtė et al., 2010) and that many older adults present sleep disturbances, with most having irregular sleep-wake patterns (Lund et al., 2010). In both cases, individual performance throughout the day seems to be affected (Schmidt et al., 2009). Obtaining sufficient sleep, and of adequate quality, is rapidly becoming a major public health concern (Colten 34 et al., 2006). The environmental and social conditions of our 24-hour societies appear to have enabled the steady and constant decline in the number of hours devoted to sleep, increasing not only deprivation, but also sleep disturbances (Cappuccio et al., 2010). The detrimental effects of this reality were first acknowledged by the industry (e.g. airlines, long-distance driving, shift-work manufacturing industry, emergency services), and later by the population at-large. This lack of rhythmicity in modern societies hinders the adaptive mechanism provided by regularity in lifestyle. The rhythmicity in lifestyle seems to aid in the maintenance of good health and well-being. Nonetheless, few studies provide solid epidemiological data on sleep-wake patterns and its evolution through the time (Youngstedt et al., 2016). Epidemiological studies on sleep are also scarce for the Portuguese population, especially for community-dwellers. Thus, it is important to work on this, so that we are able to determine the trend of evolution of sleep parameters throughout time and across ages, and understand how these correlates associate with other aspects of the individual life. 1.9. Sleep quality and its measurement There are several different methods of assessing sleep. As mentioned, the gold standard is polysomnography (PSG). However, sleep can also be estimated using diaries and wrist actigraphy. Regarding the latter, it involves wearing a wristwatch-like device that counts wrist movements (no movement counts equals sleep) and simply identifies movement versus no movement that is then converted into sleep versus wake. Compared to PSG, actigraphy has the advantage of being easily used to monitor not only for 24-hours, but also for multiple days, providing a measure of habitual behavior (Landry, Best, & Liu-Ambrose, 2015; A. Sadeh, Sharkey, & Carskadon, 1994; Avi Sadeh, 2011; Avi Sadeh & Acebo, 2002). A large set of devices is available and the choice of what to wear depends on the study purposes. On this, Table 1 provides an overview of the most used devices in research and some of their most important features. The amount of time participants wear the monitor varies across studies, but usually it is accepted that a minimum of 7-days is enough to provide a representative sample (e.g. Rowe et al., 2008). It is also common practice to ask participants to wear the device in the non-dominant wrist in order to reduce noise. Total time in bed, total sleep time, sleep latency, sleep efficiency, wake after sleep onset, number of awakenings and time of each awakening are some of the possible measures that the device can provide. It is 35 important to be aware of a number of pitfalls of actigraphy. Specifically, validity has neither been established for all scoring algorithms or devices, nor for all clinical groups, albeit it exists for some. Furthermore, actigraphy is not sufficient for the diagnosis of sleep disorders in individuals with motor disorders, or high motility during sleep and the use of computer scoring algorithms without controlling for potential artifacts can lead to inaccurate and misleading results (Sadeh and Acebo, 2002). Of note, for our studies, it was used ActiSleep+, firmware2.2.1 (ActiGraph, LLC, Pensacola, Florida, USA), a small (4.6×3.3 ×1.5 cm), electronic, light weight (19 grams), water proof, tri-axial wrist-worn device, which measure activity “counts” and are initialized at a sample rate of 30 Hz to record activities for free-living conditions. The obtained information is downloadable using ActiLife 6 software (v 6.9.0; ActiGraph, LLC, Pensacola, FL, USA) and integrated into 60-s epochs for posterior analysis using Cole-Kripke algorithm (Cole et al., 1992). While for sleep parameters, such as sleep quantity or latency, it is easy to extract information, for sleep quality the same does not occur. In fact, a clear definition of what is sleep quality has not yet been provided, which consequently lead to different studies conceiving it differently. For example, while in some studies, sleep quality is a multidimensional concept (Buysse, Reynolds, Monk, Berman, & Kupfer, 1989), in others it can either represent the adequate values of each of the different dimensions or be a global score that represents how satisfied the individual is with his/her sleep (e.g. Sun et al., 2018; Takeuchi et al., 2018). This high variability in the approach used makes it difficult to compare studies’ results. Thus, the first step should be to properly define sleep quality, so that it can be easier to compare results across studies. It is also important to state that, with a proper definition, the comparability across studies can be potentially attained not only via an objective/quantitative method, such as actigraphy or polysomnography, but also subjectively through self-report. As long as the concept measured is the same, all these methods provide for complementary information. In the Chapters II and III, this issue will be addressed in more detail. 36 Table 1. Actigraphy monitors comparison 2 . Company Actigraph Phillips Respironics CamNtech Inc SOMNOmedics America Inc Ambulatory Monitoring Inc Monitor CentrePoint Insight Watch ActiGraph GT9X Link ActiGraph wGT3XBT wActiSleep-BT Monitor3 ActiSleep+ Monitor4 Actiwatch Spectrum PRO Actiwatch Spectrum Motionwatch-8 PRO-Diary SOMNOwatch plus SOMNOwatch Micro Motionlogger Watch Motionlogger Watch Website www.actigraphcorp.com www.actigraphy.com www.camntech.com www.somnomedics.com www.ambulatory-monitoring. com Peer Reviewed Validation Articles www.actigraphcor p.com/category/ researchdatabase/sleep Cellini N, Buman MP, McDevitt EA, Ricker AA, Mednick SC. Direct comparison of two actigraphy devices with polysomnographi cally recorded naps in healthy young adults. Chronobiol Int. 2013;30(5):6918. https://onlinelibr ary.wiley.com/doi /abs/10.1111/s br.12103 https://journals.p los.org/plosone/ article?id=10.137 1/journal.pone.0 172535 https://www.ncbi .nlm.nih.gov/pmc /articles/PMC57 79275/ ---------- ---------- www.actigraphy.respironics.co m/webliography https://academic.oup.com/sle ep/article/39/6/1219/24539 57 Stevens A, et al. The effect of sleep disturbance during pregnancy and perinatal period on postpartum psychopathology in women with bipolar disorder. J Women’s Health Care. 2014;3(6):196. Rumbold PL, Doddreynolds CJ, Stevenson E. Agreement between paper and pen visual analogue scales and a wristwatch-based electronic appetite rating system (PRODiary©), for continuous monitoring of freeliving subjective appetite sensations in 7-10 year old children. Appetite. 2013; 69:180-5. ---------- ---------- Five peer-reviewed articles validate this actigraph. Example: Cole R, et al. Automatic sleep/wake identification from wrist activity. Sleep. 1992;15(5):461-9. https://www.ncbi.nl m.nih.gov/pubmed/ 25687438 Rupp TL, Balkin TJ. Comparison of Motionlogger Watch and Actiwatch actigraphs to polysomnography for sleep/wake estimation in healthy young adults. Behav Res Methods. 2011 Dec; 43(4):115260. Dimensions (cm) 4.83x3.43x1.04 3.5x3.5x1.5 4.6x3.3x1.5 4.6x3.3x1.9 4.5x3.4x1.9 4.8x3.7x1.5 4.8x3.7x1.4 3.81x2.54x1.016 5.1x3.4x0.8 4.5x4.5x1.6 4.5x4.5x1.6 3.6x3.6x1.2 5.5x4.5x1.8 Weight (g) 35 38 19 22 22 31 30 16.8 16 30 30 30 65 Time/Date Display Yes Yes No No No Yes Yes No Not reported No No Yes Yes Event Marker Button Yes No No No No Yes Yes Yes Not reported Yes Yes Yes Yes Sleep Efficiency Calculation Yes No Yes Yes Yes Not found Not found Yes Not reported Yes Yes Yes Unkown 2 Adapted from http://www.sleepreviewmag.com/2018/12/actigraphy-guide/. The website of each company was also consulted. 3 Descontinued. 4 Descontinued. 37 Sleep Latency Calculation Yes No Yes Yes Yes Not found Not found Yes Not reported Yes Yes Yes Unkown Temperature No No No No No Not found Not found No No No No Yes Yes Other Total sleep time (TST), wake after sleep onset (WASO), daytime activity (energy expenditure, steps taken, activity intensity, sedentary time), raw acceleration data Real-time data uploads to cloudbased CentrePoint software platform via home data hub and mobile application. Slim and compact design, interchangeable wrist bands. Gyroscope and magnetometer sensors for advanced positional data capture; integrated wear time sensor for off-wrist detection; support for heart rate (wireless HR sensor required); automatic bedtime detection; daytime activity profile. Compatible mobile app supports real-time data uploads and subject feedback. ---------- Integrated wear time sensor for off wrist detection, support for heart rate (wireless HR sensor required), PLM scoring, automatic bedtime detection, body position, and daytime activity profile. Compatible mobile app for iPhone, iPad, and Android platforms supports real time device communicatio n and data reports. ---------- The Actiwatch Spectrum PRO incorporates all of the features of the Actiwatch Spectrum Plus and provides subjective scoring capabilities and audible and vibrational alarms. The alarms remind subjects to enter subjective scores on a preprogrammed schedule or on a manual basis. This capability adds another dimension to data collection when studying parameters such as pain and fatigue. ---------- It is possible to collect information on circadian rhythm and 40 sleep parameters Paperless diary with OLED screen interface and touch sensitive slider. Uses the Motionware software for sleep analysis and has additional software for user designed questionnaires with 8 different question types. FDA approved and validated against paper diaries in clinical work with timed, random, or user activated prompts. Allows up to 250 questions. Questionnaires can be designed using the PRO-Diary software and then loaded via USB. Questions can be asked at given times, random times or can be user initiated with flexible scheduling of up to 30 days. Full visibility of when your subject is answering questions and how long it takes them to do so. PLMS scoring; programmable start and stop periods for several recording periods PLM scoring, 1 external channel possible (2nd Actisensor, ECG, EEG, respiratory), programmable start and stop periods for several recording periods, linking of several SOMNOwatch recordings thanks to high synchronizatio n rate. Ambient light, visual status indicator, multimode data collection, off-wrist detection. Off-wrist detection channel, PVT test, user rating scale entry, alarms (1 user and up to 10 fixable), stopwatch, countdown timer Battery Life * 30 days 14 days (wireless & gyro disabled, 30 Hz sample rate) 25 days (Wireless disabled, 30 Hz sample rate) 30 days 30 days 50 days 8 months with continuous use 90 (light sensor on), 120 (without light sensor) 14 days, assuming 10 minutes of questionnaire interface time per day. If motion logging is not selected, the PRODiary can record for up to 28 days, 30 days +26 days + 30 days + 30 days 38 assuming 10 minutes of interface time per day. Battery Options Rechargeable lithium polymer Rechargeable lithium polymer Rechargeable lithium polymer Rechargeable lithium polymer Rechargeable lithium polymer CLB 2032 lithium ion rechargeable (factory replaced) CR 2430 Lithium Coin Cell (factory replaced) Common watch CR2032 Lithium-ion rechargeable through a USB port or independent USB charger Up to 25 day study duration (depending on number of sensors attached) Lithium-ionAccu, inbuilt, rechargeable Lithium battery 1 DL2450 disposable coin cell Memory Size 4 GB 180 days/4 GB 4 GB 4 GB 4 GB 32MB Non-volatile 1 Mbits 4MB 4 MB 64MB 8 MB internal storage card 2 MB 2 MB Sample Rate 32-256 Hertz 30-100 Hertz 30-100 Hertz 30-100 Hertz 30-100 Hertz Not reported 32 Hertz 1, 2, 5, 10, 15, 30, 60 Hertz Not reported 256/32 Hertz 1 per second up to 256 per second 16 Herzt Not reported Acceleromet er Technology Primary accelerometer (±8G) 3-axis solid state accelerometer with digital filtering 3-axis solid state accelerometer with digital filtering 3-axis solid state accelerometer with digital filtering 3-axis solid state accelerometer with digital filtering MEMS type acceleromete Solid-state "Piezoelectric" acceleromet er triaxial accelerometer tri-axial accelerometer 3 activity sensors (x, y, z-axis, magnitude), ambient light, patient marker with acoustic tone 3 activity sensors (x, y, z-axis, magnitude), ambient light, patient marker with acoustic tone Zero Crossing (ZC) or Proportional Integrating Measure (PIM, selectable high or low sensitivity Solid state triaxial Water Resistance water resistant, IP57 I meter, 30 min Yes Yes Yes Yes Waterproof at 1m for 30 min per IP27 IEC 60529 Waterproof 1 m for 30 min per IPX7 IEC 60529 waterproof to 3 bar; acceptable for swimming No Not reported Not reported Yes 50 m waterresistant Light Sensor Wavelength Range (nm) n/a Yes Yes Yes Yes Yes 400 - 700 nm Three color light sensors that provide irradiance and luminous flux recordings in three color bands of the visible spectrum: red, green, and blue. wide spectrum visible Not reported n/a n/a Yes Photodiode, 400 ton 700 (520 peak) *days, during regular use 39 Objectives Sleep is a multidimensional concept that is influenced by the environment and the diversity of activities, (social) roles and contexts of the individual. Beyond the social, cultural and psychological influences, studies have shown that throughout the lifespan sleep changes are also due to biological transformations. Among these alterations are changes in both sleep patterns and routines, and in the amount of complaints reported by the individual, which seems to increase in middle-agers and older adults. Despite the marked progress in answering the fundamental questions – what is sleep, what are its mechanisms and functions –, subjective measures of sleep are still not fully understood and sleep diagnostic and monitoring tools and instruments are still evolving. Thus, herein, we propose to first dissect the meaning of a “good sleep quality” across the adult lifespan, and, then, explore associations between subjective sleep quality and neuropsychological and neuroimaging correlates. For this purpose, individuals with ages comprised between 18 and 87 years were recruited from familyand community-based health care centers, following a cross-sectional and a longitudinal design. We hypothesized that the way individuals conceive their sleep is associated to personal and psychological traits, which further associates with their rating of subjective sleep quality. Furthermore, it is suggested that functional and structural brain connectivity information might help explain, and be explained, by subjective sleep quality parameters. The results of the present work are expected to contribute to a future development of individual profiles that will allow better anticipating, predicting and acting on different health and cognitive outcomes across the lifespan. Specifically, the research questions of the present thesis are: 1. Across the adult lifespan, how do community-dwellers regard the meaning of “good sleep quality”? 2. Across the adult lifespan, are there age or sex differences in the way individuals define good sleep quality? 3. What is the association between self-rating quality of sleep and the standard measure PSQI (Pittsburgh Sleep Quality Scale)? 4. What are the determinants of “good” and “poor” sleep quality? 5. What is the association between subjective sleep quality measured by PSQI and brain functional and structural connectivity as assessed by MRI? 40 We hypothesized that: 1. Individuals conceive sleep quality as a multidimensional concept and the parameters reported as important for a good sleep quality might change with age and sex; 2. There is an association between self-rating scales and the standard PSQI. However, considering that different people might define a good sleep quality differently, we expect this association to be moderate. We also speculate that the type of scale used to determine sleep quality influences the amount of bias on its scoring; 3. Social, personal and psychological variables are known to influence sleep. Thus, variables such as age, sex, marital status, psychological morbidity, and certain personality traits, associate with subjective sleep quality; 4. Functional connectivity (FC) decreases with poor sleep quality. Given that previous studies have shown that decreases in synchrony occur with sleep loss and fragmentation, we expect that this can also be observed with a subjective measure of sleep quality. Furthermore, because this measure reports to the previous month, it is also expected that it will be sufficient to allow to observe changes in structural connectivity (SC). As such we propose to: 1) review the construct of sleep quality (Chapter II); 2) study how adult community-dwellers across the lifespan conceive “good sleep quality” (Chapter III); 3) characterize sleep patterns and routines (Chapter III and Chapter IV); and 4) determine the associations between subjective sleep quality and psychological variables and brain correlates (Chapter III and Chapter V). Ultimately, the work is expected to lay a foundation for raising awareness about misadjusted sleep/rest behaviors and/or patterns, and develop and promote strategies that empower the individual to make better and healthier decisions. 47 Ohayon, M. M., & Paiva, T. (2005). Global sleep dissatisfaction for the assessment of insomnia severity in the general population of Portugal. Sleep Medicine , 6 (5), 435–441. https://doi.org/10.1016/j.sleep.2005.03.006 Paixão, E., Branco, M. J., & Contreiras, T. (2006). Uma observação sobre a prevalência de perturbações do sono, em Portugal Continental (report). Instituto Nacional de Saúde Doutor Ricardo Jorge. Observatório Nacional de Saúde. Obtido de http://repositorio.insa.pt/handle/10400.18/283 Paranjpe, D., & Sharma, V. (2005). Evolution of temporal order in living organisms. Journal of Circadian Rhythms , 3 (0). https://doi.org/10.1186/1740-3391-3-7 Payne, J. D., Schacter, D. L., Propper, R. E., Huang, L.-W., Wamsley, E. J., Tucker, M. A., … Stickgold, R. (2009). The role of sleep in false memory formation. Neurobiology of Learning and Memory , 92 (3), 327–334. https://doi.org/10.1016/j.nlm.2009.03.007 Pevet, P., & Challet, E. (2011). Melatonin: Both master clock output and internal time-giver in the circadian clocks network. Journal of Physiology-Paris , 105 (4–6), 170–182. https://doi.org/10.1016/j.jphysparis.2011.07.001 Pilcher, J. J., Ginter, D. R., & Sadowsky, B. (1997). Sleep quality versus sleep quantity: Relationships between sleep and measures of health, well-being and sleepiness in college students. Journal of Psychosomatic Research , 42 (6), 583–596. https://doi.org/10.1016/S0022-3999(97)00004-4 Rittman, M., Hinojosa, M. S., & Findley, K. (2009). Subjective sleep, burden, depression, and general health among caregivers of veterans poststroke. The Journal of Neuroscience Nursing: Journal of the American Association of Neuroscience Nurses , 41 (1), 39–52. Sadeh, A., Sharkey, K. M., & Carskadon, M. A. (1994). Activity-based sleep-wake identification: an empirical test of methodological issues. Sleep , 17 (3), 201–207. Sadeh, Avi. (2011). The role and validity of actigraphy in sleep medicine: An update. Sleep Medicine Reviews , 15 (4), 259–267. https://doi.org/10.1016/j.smrv.2010.10.001 Sadeh, Avi, & Acebo, C. (2002). The role of actigraphy in sleep medicine. Sleep Medicine Reviews , 6 (2), 113–124. Saper, C. B., & Fuller, P. M. (2017). Wake-sleep circuitry: an overview. Current Opinion in Neurobiology , 44 , 186–192. https://doi.org/10.1016/j.conb.2017.03.021 48 Saper, C. B., Fuller, P. M., Pedersen, N. P., Lu, J., & Scammell, T. E. (2010). Sleep State Switching. Neuron , 68 (6), 1023–1042. https://doi.org/10.1016/j.neuron.2010.11.032 Saper, C. B., Scammell, T. E., & Lu, J. (2005). Hypothalamic regulation of sleep and circadian rhythms. Nature , 437 (7063), 1257–1263. https://doi.org/10.1038/nature04284 Schabus, M., Dang-Vu, T. T., Albouy, G., Balteau, E., Boly, M., Carrier, J., … Maquet, P. (2007). Hemodynamic cerebral correlates of sleep spindles during human non-rapid eye movement sleep. Proceedings of the National Academy of Sciences of the United States of America , 104 (32), 13164–13169. https://doi.org/10.1073/pnas.0703084104 Scott, C., & Sherrington, S. (2009, Julho). Man on his Nature by Charles Scott, Sir Sherrington. https://doi.org/10.1017/CBO9780511694196 Sharkey, K. M., & Van Reen, E. (2014). The “Realities” of our Modern Light-Dark Cycle. Journal of Clinical Sleep Medicine : JCSM : Official Publication of the American Academy of Sleep Medicine , 10 (7), 723–724. https://doi.org/10.5664/jcsm.3856 Sharma, V. K. (2003). Adaptive significance of circadian clocks. Chronobiology International , 20 (6), 901–919. Shokri-Kojori, E., Wang, G.-J., Wiers, C. E., Demiral, S. B., Guo, M., Kim, S. W., … Volkow, N. D. (2018). β-Amyloid accumulation in the human brain after one night of sleep deprivation. Proceedings of the National Academy of Sciences , 201721694. https://doi.org/10.1073/pnas.1721694115 Siegel, J. M. (2009). Sleep viewed as a state of adaptive inactivity. Nature Reviews. Neuroscience , 10 (10), 747–753. https://doi.org/10.1038/nrn2697 Simpson, C., & Carter, P. A. (2010). Pilot study of a brief behavioral sleep intervention for caregivers of individuals with dementia. Research in Gerontological Nursing , 3 (1), 19–29. https://doi.org/10.3928/19404921-20090731-02 Smedje, H., Broman, J.-E., & Hetta, J. (1998). Sleep disturbances in Swedish pre-school children and their parents. Nordic Journal of Psychiatry , 52 (1), 59–67. https://doi.org/10.1080/080394898422580 Song, Y., McCurry, S. M., Fung, C. H., Josephson, K. R., Teng, E., Irwin, M. R., … Martin, J. L. (2018). Sleep and Caregiving Experiences Among Caregivers of Veterans in an Adult Day Health Care Program: A Pilot Study. Clinical gerontologist , 41 (2), 167–171. https://doi.org/10.1080/07317115.2017.1365795 49 Stevner, A. B. A., Vidaurre, D., Cabral, J., Rapuano, K., Nielsen, S. F. V., Tagliazucchi, E., … Kringelbach, M. L. (2019). Discovery of key whole-brain transitions and dynamics during human wakefulness and non-REM sleep. Nature Communications , 10 (1), 1035. https://doi.org/10.1038/s41467-019-08934-3 Stickgold, R. (2006). Neuroscience: a memory boost while you sleep. Nature , 444 (7119), 559– 560. https://doi.org/10.1038/nature05309 Stickgold, R., & Walker, M. P. (2010). The Neuroscience of Sleep . Academic Press. Suntsova, N., Szymusiak, R., Alam, N. M., Guzman‐Marin, R., & McGinty, D. (2002). Sleep-waking discharge patterns of median preoptic nucleus neurons in rats. The Journal of Physiology , 543 (2), 665–677. https://doi.org/10.1113/jphysiol.2002.023085 Tononi, G., & Cirelli, C. (2006). Sleep function and synaptic homeostasis. Sleep Medicine Reviews , 10 (1), 49–62. https://doi.org/10.1016/j.smrv.2005.05.002 Troxel, W. M., Buysse, D. J., Matthews, K. A., Kravitz, H. M., Bromberger, J. T., Sowers, M., & Hall, M. H. (2010). Marital/Cohabitation Status and History in Relation to Sleep in Midlife Women. Sleep , 33 (7), 973–981. Virtanen, M., Ferrie, J. E., Gimeno, D., Vahtera, J., Elovainio, M., Singh-Manoux, A., … Kivimäki, M. (2009). Long Working Hours and Sleep Disturbances: The Whitehall II Prospective Cohort Study. Sleep , 32 (6), 737–745. von Känel, R., Mausbach, B. T., Ancoli-Israel, S., Dimsdale, J. E., Mills, P. J., Patterson, T. L., … Grant, I. (2012). Sleep in spousal Alzheimer caregivers: a longitudinal study with a focus on the effects of major patient transitions on sleep. Sleep , 35 (2), 247–255. https://doi.org/10.5665/sleep.1632 Walker, M. P., & Stickgold, R. (2004). Sleep-dependent learning and memory consolidation. Neuron , 44 (1), 121–133. https://doi.org/10.1016/j.neuron.2004.08.031 Williams, S., Meadows, R., & Arber, S. (2010). The sociology of sleep (pp. 275–299). Wulff, K., Gatti, S., Wettstein, J. G., & Foster, R. G. (2010). Sleep and circadian rhythm disruption in psychiatric and neurodegenerative disease. Nature Reviews. Neuroscience , 11 (8), 589– 599. https://doi.org/10.1038/nrn2868 Xie, L., Kang, H., Xu, Q., Chen, M. J., Liao, Y., Thiyagarajan, M., … Nedergaard, M. (2013). Sleep Drives Metabolite Clearance from the Adult Brain. Science (New York, N.Y.) , 342 (6156). https://doi.org/10.1126/science.1241224 50 Youngstedt, S. D., Goff, E. E., Reynolds, A. M., Kripke, D. F., Irwin, M. R., Bootzin, R. R., … JeanLouis, G. (2016). Has Adult Sleep Duration Declined Over the Last 50+ Years? Sleep medicine reviews , 28 , 69–85. https://doi.org/10.1016/j.smrv.2015.08.004 51 CHAPTER II The importance of defining what we are measuring: a systematic review on sleep quality meaning Liliana Amorim, Nuno Sousa, Nadine Correia Santos (Submitted) 52 The importance of defining what we are measuring: a systematic review on sleep quality meaning Authors: Liliana Amorim,2,3, Nuno Sousa1,2,3, Nadine Correia Santos1,2,3* 1Life and Health Sciences Research Institute (ICVS), School of Medicine, University of Minho, Campus Gualtar, Braga, Portugal 2ICVS/3B’s - PT Government Associate Laboratory, Braga/Guimarães, Portugal. 3Clinical Academic Center – Braga (2CA-B), Braga, Portugal. *Corresponding author: Nadine Correia Santos, Life and Health Sciences Research Institute (ICVS), School of Medicine, University of Minho, Campus Gualtar, 4710-057 Braga, Portugal. Email: [email protected]. Phone: +351 253 604 806. 53 Abstract To date (December 2018), there are approximately 18 000 articles on PubMed addressing sleep quality. Of these, 2256 just in the last one-year period. However, the vague use of terminology may be contributing to the variability of study results. Here, to understand the state of the art, and clarify the meaning of the concept ‘sleep quality‘, we performed a systematic review on what is sleep quality and how can it be operationalized. Studies considering for a definition of sleep quality, subjective sleep or perceived sleep, were included. Results show that few studies indicate which elements should be considered in defining sleep quality, with none providing a definition, albeit, some indicating parameters that contribute for it. More so, studies are not uniform as to the measures used, resulting in different interpretations of the sleep quality construct among the general population. Older studies, combining subjective and objective measures, and promoting variability within the study design, considered sleep continuity variables as the most important element contributing for subjective sleep quality construct. However, more recent studies focusing only on qualitative data seem to indicate that the main feature when reporting a good sleep quality relates to next day performance and the memory of the previous night sleep. Given the importance of ‘sleep quality‘ construct not only for research, but also for clinical practice, namely in its monitoring and in diagnosis and treatment of sleep-related disorders, it is of crucial importance to clarify its meaning and definition. Key-words: sleep quality; definition; meaning; construct; subjective sleep quality; systematic review 54 Introduction Sleep is a multidimensional construct, often referred to as in terms of its quality (Ohayon et al., 2017). Indeed, sleep quality has been a well-recognized predictor of physical and mental health, cognitive status, wellness and overall vitality (Alhola & Polo-Kantola, 2007; Asif, Iqbal, & Nazir, 2017; Besedovsky, Lange, & Born, 2012; Freeman et al., 2017; Lim & Dinges, 2010; Ohayon et al., 2017; Simon & Walker, 2018; L. Xie et al., 2013). However, among the results of populationbased studies, there is a concerning heterogeneity which might be associated to different conceptualizations of ‘sleep quality‘. Efforts to define sleep quality have been mainly driven by clinical considerations and may not necessarily reflect the general population understanding of it (Goelema et al., 2018). This is particularly relevant because often, it is the perception of poor sleep quality that takes the individuals to seek their doctor (Akerstedt, Hume, Minors, & Waterhouse, 1997). Furthermore, physicians frequently rely on objective measures of sleep to diagnose sleep disturbances, which has been shown in the literature not to correlate to self-reports of poor sleep quality. For instance, studies indicate that objective sleep parameters, driven from polysomnographic (PSG) or EEG measurements, often do not translate self-reported measures of poor sleep quality or a history of chronic insomnia (Moul et al., 2002; Rosa & Bonnet, 2000). Adding to this, there is a lack of operationalization of the concept ‘sleep quality‘ or, at least, an heterogeneity in the parameters that are being used to refer to it (e.g. Kwok et al., 2018). For example, while some studies use ‘sleep satisfaction‘ (e.g. Delaney et al., 2018) or ‘depth of sleep‘ (e.g. Takeuchi et al., 2018) when addressing ‘sleep quality‘, others use self-ratings of the previous night (e.g. Vitale et al., 2018) or the Pittsburgh Sleep Quality Index (PSQI) (e.g. Li et al., 2017). Indeed, this is an issue already identified by other authors (e.g. Harvey et al., 2008) that also suggest that the vague use of terminology might be a possible reason for the variability among study results. To understand the state of the art on this topic, and to determine what should be the future steps for providing an objective and adequate definition of Sleep Quality, we performed a systematic review on what is ‘sleep quality‘ and how it can be operationalized. 55 Methods Literature search on the measures used for sleep quality assessment The first step was to perform a PubMed search with the expression ‘sleep quality‘. The purpose of this was to determine the tools that are being used for ‘sleep quality‘ assessment and, consequently, to understand how this concept is been conceptualized, i.e., the choice of the instrument used has implicit a particular definition of the construct. The search was conducted on December 4th (2018) and reviewed on December 13rd. Due to the expected high number of articles, results were restricted to the first 100 articles and to last one-year publications. Articles were then screened regarding the measure/tools used for sleep quality assessment. Systematic review of the subjective meaning of a good sleep quality/a good night of sleep Next, a systematic review addressing the question of how the general adult population (18 years and more) conceived the meaning of a good sleep quality was conducted. Different search engines were considered and the search expression was adjusted to its specificities. The expression used for the search on PubMed was ("subjective sleep"[All Fields] OR "perceived sleep"[All Fields] OR "sleep quality"[All Fields]) AND (meaning[All Fields] OR definition[All Fields]) , sorted by Best Mach . The search was performed on December 4th (2018) and reviewed on January 10th (2019), in order to guarantee that no new results would be left out of the analysis. No date, language or article type restrictions were imposed to the search. To guarantee that all literature of interest was considered, a search in EBSCO (through Psychology and Behavioral Sciences Collection) was also performed. The search expression used was ("subjective sleep quality meaning" OR "sleep quality definition" OR ("perceived sleep" AND "meaning")). A search was also performed in Google Scholar with the search expression ("subjective sleep quality meaning" OR "sleep quality definition"), to guarantee that no potential articles of interest were left without consideration. A cross-ref was also executed. All articles that aimed to clarify or contribute to the meaning/definition of a ‘good sleep quality‘ were considered, including reviews or works that used objective measures of sleep quality, as long as it was in association to the meaning of this concept. All articles that filled the inclusion criteria were considered for the literature analysis (Figure 1). 56 Figure 1. Flow diagram of the systematic review process. Results Heterogeneity of the measures used to assess sleep quality From the first part of the work (the PubMed search on the expression ‘sleep quality‘ for the determination of the tools being used for sleep quality assessment), a total of 17674 results were obtained. When restricting the results to last one-year (January to December 2018), 2253 articles remained. From these, in order to determine the measures being used to assess sleep quality, and how it was being defined, the first 100 articles were included for further analysis (Table 1). Of Records identified through PubMed search (n = 82) Screening Included Eligibility Identification Records identified through EBSCO search (n = 2433) Records after duplicates removed (n = 2502) Records screened by title (n = 5513) Records excluded (n = 3789) Abstracts assessed for eligibility (n = 1724) Abstracts excluded (n = 1486) Full-text articles assessed for eligibility (n = 238) Studies included in qualitative synthesis (n = 13) Records identified through Google Scholar search (n = 5500) Full-text excluded (n = 225) 63 Back to basics: a literature review on sleep quality meaning and definitions Regarding the systematic review addressing the meaning and/or definition of sleep quality, a total of 238 articles were considered for full text analysis (Figure 1). Of these, 13 were included in the systematic review, based on the defined criteria. Table 2 presents a simple characterization of the included studies regarding the country where it was performed, the parameter(s) or definition of sleep quality used in the study and the most relevant limitations. In Table 3, a detailed characterization of the main aspects of each of these studies is provided in order to better elucidate the reader. Table 2. Summary table of the principal limitations of the studies considered for the systematic review and indication of the country in which the information was collected. Reference Data Collection Place Sleep quality definition Study limitations (Goelema et al., 2018) The Netherlands Daytime functioning as the more important parameter Results from literature review are not shown. Small sample size for the age range considered. (Ramlee et al., 2017) England Mostly influenced by memories of what happened during sleep and their experience upon waking Small sample size. Only young individuals were considered. (Ramlee et al., 2018) England Is not solely determined by nighttime parameters but also by daytime processes through retrospective judgment Small sample size. Only young individuals were considered. (Rosipal et al., 2013) European Objective and subjective sleep parameters and its association witht daytime quality variables (n.a. – review) (Krystal & Edinger, 2008) USA Likert-style rating of the previous night’s sleep quality. Sleep quality is only considered as a score from a rating of the previous nigh; (Harvey et al., 2008) UK Subjective feelings regarding the day following sleep appeared to be the most important basis for judging sleep quality Small sample size. Young individuals only. (Yi et al., 2006) South Korea Sleep quality was defined considering the answers to the questions: How is your sleep these days? What do you think a good sleep is? What do you think of a poor sleep? Small sample size for all the groups considered; The same population was used to develop and test a new instrument to assess sleep quality. (Keklund & Akerstedt, 1997) Sweden Subjective sleep quality is related to perceptions of ease of initiation and maintenance of sleep. From a physiological point of view, subjective sleep quality seems to be a matter of SWS and sleep continuity (i.e. as indicated by sleep efficiency). Small sample size. Only young individuals were considered. (Akerstedt et al., 1994) Sweden Mainly involved variables of sleep continuity, in particular, perceived calmness of sleep and sleep efficiency. The same sample is used in both studies. Small sample size. Mostly young individuals. In one of the studies only females were considered. (Åkerstedt et al., 1994) Sweden Mainly related to sleep efficiency but also to the closeness of the awakening to the circadian acrophase (Buysse et al., 1989) USA Composite measure of different parameters, namely sleep duration, latency, disturbance, medication, day dysfunction, perceived quality. Developed PSQI Index was defined only considering clinical implications. 64 (Snyder‐Halpern & Verran, 1987) USA Described in terms of sleep fragmentation (i.e., midsleep awakenings and movement during sleep); sleep length (i.e. the total sleep period); delay (i.e. sleep latency) and depth (i.e. soundness of sleep, rest upon awakening; method of awakening; subjective quality of sleep. It was only considered a clinical population; There is no comparison of this measure with other standard methods. (Johns et al., 1971) Australia Definition derived from two questions: at what time do you usually go to bed at night on weekdays? how would you describe your usual sleep? Small sample size Only young individuals and students are being considered, Overall, few studies (5.56%) addressed the meaning of “sleep quality” in the general population. Despite in most studies authors mentioned that a review of the literature was performed, none of them presented the results of such review nor its implications for their approach to the problematic. Studies were comprised of small samples (ranging in sample size from n= 16 to n= 64), two employed the same sample (Åkerstedt et al., 1994; Akerstedt et al., 1994), and most (61.5%) recurred to young college students. A comparison between “normal sleepers” (defined as individuals without sleep complaints or reported poor sleep quality) and “problematic populations” (such as individuals with chronic pain or sleep disturbances) was also addressed in some of the studies (Ramlee et al., 2018). In terms of findings, it was possible to observe an association between sleep quality and the (reported) delay to sleep, total time awake, duration of night sleep, and nightmares (Akerstedt et al., 1994; Goelema et al., 2018; Ramlee et al., 2018, 2017). A "good sleep quality" was mainly reported as a question of sleep continuity variables (Åkerstedt et al., 1994; Akerstedt et al., 1994). However, in the most recent studies (Goelema et al., 2018; Ramlee et al., 2018, 2017), this is no longer the case as sleep quality judgment seems to depend more on matters of performance in the next day, which is aligned to the memory that the individual has of the previous sleep time. Briefly, we observe that: (1) studies developed to understand how sleep quality is interpreted by the general population are few; (2) except for one study (Goelema et al., 2018), all studies considered only samples with young individuals, including those that explored for comparisons between insomniacs vs good/normal sleepers; (3) at no moment it is used an approach in which is the participant had to think and present the parameters that he/she uses to define a good sleep quality or a good night of sleep; (4) there was no attempt to determine the association between the reported parameters of subjective sleep quality and existing standard measures, such as PSQI; 5) no determination of the association of the profiling of answers and other psychological parameters, or even sleep routines, was performed. 65 Table 3. Summary table of the information from the studies that also explored the meaning of “sleep quality”. Year Journal Author Country Title Aims Methods Results Conclusions 2018 BMC Research Notes (Goelema et al., 2018) Netherlands Conceptions of sleep experience: a layman perspective To understand the essence of the sleep experience and the concepts held by lay people without sleep disorders Cross-sectional study. n=64 respondents 32 females; age range: 18-79yrs Young (≤ 49 years)/Old (≥ 50 years) Study conducted online. 4 participants per group of sex, age (young and old), education (high, ≥ bachelor; low, < bachelor) and PSQI (good, ≤ 5; poor, > 5). Sample of healthy individuals – not actively screened for signs of undiagnosed-sleep disorders. Sleep Sentence Completion Questionnaire (SSCQ) (projective data collection technique) was used to survey participants’ conceptions of sleep experience. This measure was obtained after a pilot study in which 10 participants answer to a sample of a 62-item stem completion questionnaire that was then analyzed in terms of word frequency. Demographic information, educational level and the Dutch version of the Pittsburgh Sleep Quality Index (PSQI) were also applied. Objective sleep measurements not performed. A phenomenological data analysis approach was adopted. Two coders performed independently a direct content analysis to ensure the validity of the clustering. The two coders achieved a consistency of 0.81 Cohen’s kappa. PSQI range: 1-16 Bed time_ 23:38:59 (1:13:42) Wake up time_ 7:31:41 (1:11:35) 9 themes resulted from data analysis: next day state; interruptions during the night; before bed state; sleep characteristics; bedroom environment; thoughts about sleep; routine; alarm clock; other. The largest category in the analysis was ‘next day state’, followed by ‘interruptions during the night’ and ‘before bed state’. In the category ‘other’ were statements concerning dreams/nightmares, regular bedtimes, bad food and some few single statements. This category also involved statements about sleeping posture, sleep rhythm and sleeping on time. The experienced sleep quality is not depending solely on the progress of the night. Daytime functioning seems more important for people to judge their sleep experience than the actual night itself. Sleep quality definition, from a subjective point of view, should involve factors, such as stress/well-being levels, rest feeling and functioning during the day. This implies that the experienced sleep quality is not only depending on the progress of the night. These results can guide future research to provide suitable psychometric measures for normal sleepers, as well as the design of sleep data visualization applications in the context of health self-monitoring. 2017 Sleep (Ramlee et al., 2017) England What Sways People’s Judgment of Sleep Quality? A Quantitative ChoiceTo examine the relative weight that specific factors carry in the sleep quality judging process. Cross-sectional study. Sample characteristics n=111, age range: 18 and 30 years Recruited from a university-wide subject panel The parameters that occurred during the day before sleep did not have a significant impact on the participants’ choices (amount of activity: p = .38; day went well?: p = .93; mood: p = .19). Participants were asked to make choices between 2 concrete scenarios and indicate with their choice which scenario represents a better (or worse) night’s sleep. By conceptualizing the sleep quality judgement as a decision-making process, 66 Making Study With Good and Poor Sleepers To determine the possible interaction between the parameters of sleep quality extracted from different periods, between different types of sleeper, and between different types of judgment. Excluded individuals: - n=11, because they did not show for the experimental session; - n=7 for noncompliance to the task instruction. - n=6 due to the methodological issues. Groups: - “good sleeper”: scored 7 or below on the Insomnia Severity Index (ISI) n=44, male/female ratio: 20/24 - “poor sleeper”: scored 8 or above on the ISI and experienced 1 or more of the following symptoms for at least 3 nights a week, during at least 3 months, despite having an adequate opportunity to sleep: (1) difficulty initiating sleep (taking longer than 30 minutes to fall asleep), (2) difficulty staying asleep (frequent midnight awakenings), (3) early morning awakening with an inability to return to sleep, (4) daytime functioning impairment (e.g., poor concentration, excessive sleepiness). n=43, male/female ratio: 28/15 Questionnaires for characterization: - information about participant’s demographics - typical sleep pattern - insomnia severity in the past 3 months Experimental session - Self-report sleep quality was conceptualize as a decision-making process. - Placed in small groups of 3 to 4 participants in a lab with multiple computers partitioned into stations. The lab was sound attenuated with central air conditioning and lighting control. Each participant was assigned to a computer at some distance from the others to minimize distraction and response contamination. - The participants were asked to read and imagine themselves being the person experiencing 48 pairs of scenarios. They read a pair of scenarios in each trial and were asked to choose one scenario from Of the pre-sleep parameters, only physiological arousal (p < .001) had a significant impact on the participants’ choices (readiness to sleep: p = .06; cognitive arousal: p = .09). Of the sleep parameters, SOL (p < .001), WASO (p < .001) and TST (p < .001) had a significant impact, whereas memory of dream (p = .08) did not have a significant effect on the participants’ choices. Both of the upon waking parameters had a significant impact (feeling refreshed: p < .001; motivated to get up: p < .001). All of the day after parameters had a significant impact (alertness: p = .01; thinking: p < .001; mood: p < .001; sociability: p < .001; physical activity: p < .001). The most important individual parameter of sleep quality, was TST, followed by feeling refreshed (upon waking), then mood (day after) and then motivated to get up (day after). The most important time period was during sleep, followed by upon waking, then day after, then pre-sleep, and finally the parameters that occurred day before sleep were least important. The “best-preferred scenario” for a better night’s sleep was as follows, with words in bold/italic indicating the adjustable option of the 11 significant parameters: “I felt very comfortable lying in bed. It took me no time to fall asleep. I slept through the night . I think I slept for 9.5 hours . This morning, I felt somewhat refreshed on waking. I felt motivated to get out of bed. During the day, I felt alert the authors managed to quantitatively identify and estimate the relative importance of different sleep and non-sleep parameters in influencing their judgement of sleep quality. 11 out of 17 identified sleep quality parameters were found to have a significant effect on the participants’ sleep quality judgment. - participants relied most heavily on TST, feeling refreshed (upon waking) and mood (day after) to make their judgment of sleep quality. - participants’ judgment of sleep quality was most influenced by their memories of what happened during sleep and their experience upon waking, followed by their feelings and functioning during the day after, then pre-sleep experience of the night before, and lastly their experience the day before. Synergetic effects were found between: - WASO and feeling refreshed (upon waking); - feeling refreshed (upon waking) and types of question. However, whether the participant was a good or poor sleeper did not appear to make a difference in the way in which the sleep quality judgment was made.. 67 each pair that represents a night of better (or worse) sleep quality, depending on the question that they were presented. To avoid misunderstanding what was being expected from the tasks, in addition to verbal explanations the participants were given detailed written instructions on the computer screen. Each scenario described a self-reported experience of sleep, in the first person narrative, stringing together 17 possible determinants of sleep quality that we had identified from our literature review. The data from these trials were used to evaluate the relative importance of each determinant (sleep quality parameter). Data were analysed using the statistical software R (http://www.r-project.org/). Descriptive statistics were used to describe participants’ characteristics. Means and standard deviations were presented to describe continuous variables, whilst frequencies and percentages were reported for categorical variables. Independent sample t-test and chisquare statistics were used to describe the differences in characteristics between the good and poor sleeper groups. and my head was reasonably clear . My mood was good . I was somewhat sociable and physically I was reasonably active today”. Only WASO and feeling refreshed had a significant interaction (p < .001). This interaction judged a night with both WASO and feeling unrefreshed to be a particularly poor night’s sleep. However, if participants either felt at least somewhat refreshed or if they slept through the night, then they judged it to be a reasonably good night’s sleep. There was no significant interaction between parameters and types of sleeper. The interaction between parameters and types of question allowed to statistically test whether participants used the same parameters to define a good and a bad night’s sleep. Only one significant interaction was found between feeling refreshed and types of question (p = .003), suggesting that feeling refreshed was more important to the participants when judging a good night’s sleep than when judging a poor night’s sleep. 2016 Behavioral Sleep Medicine (Ramlee et al., 2018) England Do People With Chronic Pain Judge Their Sleep Differently? A Qualitative Study To extend the investigation of sleep quality and its definition to people with chronic pain. . Cross-sectional study. Qualitative study. Inductive qualitative approach to explore the mental representations of sleep quality in the patients’ mind. In-depth one-to-one interviews were carried out to provide the data and context for the researchers to interpret and extract meanings. Sample: n=17 Characteristics - Sex ratio: 9 male, 8 female - Mean age 42.1 ±15.5 (age range: 19 to 64 years) Four themes resulted from the thematic analysis. Specifically: - Theme 1: Memories of nighttime sleep disruptions Clear consensus that the participants judged their sleep quality based on their remembered ability to “switch off” and stay asleep. Awakenings in the middle of the nights were cited as indicators of poor sleep quality; the more memories of wakefulness, the stronger the feeling of having had a bad night’s sleep. A good Sleep quality is not solely determined by nighttime parameters but also by daytime processes through retrospective judgment. Particularly, people with chronic pain view pain experience and sleep quality as two linked entities that influence their ability to engage in daytime activities as planned. To the sleepers, using indirect indicators to infer sleep quality is only natural as they do not have access to sleep assessment technology and the experience of sleep is marked by darkness, loss of 68 - Mean BMI 27.9±5.89 - Work status: 7 were in full-time employment, 7 were on sick leave, medically retired, retired or not working, and the remaining 3 were studying fulltime. Groups: - Chronic widespread musculoskeletal pain (fibromyalgia)_ n=6 - Chronic localized musculoskeletal pain (back pain)_ n=5 - Absence of chronic pain_ n=6 healthy individuals Recruitment Participants were recruited through advertisements circulated within local pain patient support groups and flyers displayed across the university campus and the local community. All participants in the fibromyalgia and back pain groups confirmed that they had received a formal diagnosis of fibromyalgia or back pain from a physician. Inclusion criteria (a) aged between 18 and 65 years (b) English-speaking (c) for participants in the fibromyalgia or back pain group: the presence of pain for at least six months Exclusion criteria (a) physical disabilities or neurological disorders that prevent them from completing the questionnaire or attending the interview (e.g., visual impairment, dementia); (b) severe psychiatric illnesses (e.g., psychosis); (c) sleep disorders that might explain sleep disturbance (e.g., sleep apnea, narcolepsy). Questionnaires - a blank body manikin to assess the spread of pain (Lacey, Lewis, Jordan, Jinks, & Sim, 2005), night’s sleep was typically characterized by the general absence of interruptions to sleep and absence of memory of noise or any non-sleep activities. - Theme 2: Feelings on waking and cognitive functioning during the day Feeling refreshed on waking emerged as a key criterion of good quality sleep; when they felt refreshed by sleep they would be motivated to get up and be ready to start the day without any hesitation. In contrast, a poor night’s sleep was generally associated with a struggle to get up in the morning, tiredness on waking, and the desire to stay in bed and get some more sleep. The feeling of being refreshed by sleep appeared to be linked to the ability to overcome the sleep inertia upon transitioning from sleep to wakefulness. The participants also retrospectively judged their sleep quality based on their daytime task performance. They noted that a night of poor sleep was typically followed by a day of forgetfulness and mind-wandering. Theme 3: Ability to engage in daytime physical and social activity Following a poor night’s sleep, they tended to find themselves avoiding social engagements. Lacking energy, they would cancel appointments to give themselves an opportunity to catch up on sleep. Daytime fatigue and social withdrawal during the day were perceived to be indicators of poor quality sleep. Theme 4: Changes in physical symptoms and pain intensity The participants paid attention to their bodily sensations when they made consciousness, and amnesia. The current findings highlight the potential benefits of targeting daytime symptoms in attempts to improve sleep quality. 69 - the Brief Pain Inventory to examine pain severity and interference (BPI; Cleeland & Ryan, 1994), - Insomnia Severity Index to assess sleep problems (ISI; Bastien, Vallieres, & Morin, 2001), - Epworth Sleepiness Scale to measure daytime sleepiness (ESS; Johns, 1991), - Multidimensional Fatigue Inventory to assess fatigue (MFI; Smets, Garssen, Bonke, & Haes, 1995), - Hospital Anxiety and Depression Scale to assess symptoms of anxiety and depression (HADS; Zigmond & Snaith, 1983), - Dysfunctional Beliefs and Attitudes About Sleep Scale (DBAS; Morin, Vallieres, & Ivers, 2007) - several standard questions about the participants’ demographics such as age, sex, body mass index (BMI), and employment status. Semistructured interview - approximately 40 minutes long - participants were invited to talk in depth about their current sleep patterns and how they make judgments about their sleep quality - five open-ended questions were presented to ensure coverage of these topics All interviews were audio-recorded and transcribed verbatim by an independent professional transcriber. The transcripts were then reviewed by the interviewer (FR) and another member of the research team (EA) for accuracy. Data analysis A thematic analysis was carried out on all transcripts in accordance with the Braun and Clarke (2006) guidelines. Steps in data analysis: - the lead author (FR) familiarized herself with the data by reading and rereading the transcripts. Initial ideas and impressions related to the judgment of their sleep quality. Physical symptoms (e.g., headache, migraine and sore eyes) and unexpected loss of appetite were used to infer poor sleep quality. For participants with fibromyalgia or back pain, they factored in their current pain when judging sleep quality. These participants perceived an increase in pain as an indicator of poor night’s sleep and showed appreciation of the self-perpetuating cycle of pain and poor sleep. They believed that a poor night’s sleep would aggravate pain and fuel the risk of re-injury. 70 research questions were noted and highlighted. This step allowed the researcher to develop a thorough understanding of the data. - initial codes (i.e., brief description of the concepts identified from the data) were constructed as transcripts were being read again. All the coded data were then collated and semantically arranged. - potential themes were extracted from the coded data. - potential themes were carefully reviewed and, at this stage, the researcher consulted and discussed with a senior researcher with clinical and research experience in pain and sleep (NT) regarding the precision of the themes and the relevance of the coded data. Differences in opinions were resolved by discussion. - to ensure our interpretation did not deviate from original meaning of the data, the extracted themes and codes were sent to a subsample of the participants (n = 7) for validation. Feedbacks from the participants were incorporated into the final stage of analysis, which led to the naming of each theme. The coded data were arranged into a table in accordance with the themes they supported. When generating the themes, the researchers not only paid attention to words used by the participants, but also the context in which the participants articulated themselves. - the researchers compared and contrasted the themes across fibromyalgia, back pain, and the healthy groups, which allowed the researchers to examine whether people with chronic pain judged their sleep quality differently from those without chronic pain, and whether people with fibromyalgia evaluated their sleep quality differently from people with back pain. 71 2013 Biological Psycholog y (Rosipal et al., 2013) Europe In search of objective components for sleep quality indexing in normal sleep To investigate to what extent polysomnographic (PSG) recordings of nocturnal human sleep can provide information about sleep quality in terms of correlation with a set of daytime measures. Cross-sectional, multi-centric study. 2008 Sleep Medicine (Krystal & Edinger, 2008) USA Measuring sleep quality To consider objective measures of the subjective “sleep quality” experience. Review. A Likert-style rating of (the previous night’s) sleep quality was used as the core sleep quality indicator. Employed simple Likert-style rating of (the previous night’s) sleep quality, commonly included as an item on sleep diaries and used here as the core sleep quality indicator. Potential objective measures discussed include polysomnography, cyclic alternating pattern and actigraph The major factor limiting research on sleep quality is the lack of a standard definition. 2008 SLEEP (Harvey et al., 2008) UK The Subjective Meaning of Sleep Quality: A Comparison of Individuals with and without Insomnia To conduct a detailed and systematic investigation of the subjective meaning of sleep quality among individuals who meet diagnostic criteria for insomnia compared with a group of normal sleepers. To determine which sleep quality variables are judged as most important. To use a qualitative approach to determine whether there are important variables influencing perception of Cross-sectional. Comparisons between groups: insomniac (n = 25) and normal sleepers (n = 28). Participants’ recruitment: - from January to July, 2004; - via flyers posted around the city and referrals from primary care physicians. - the sample was a non–treatment seeking sample that was drawn from a university city and included university students. Inicial sample: n = 208; Final Sample: n = 53 Exclusions: Total_ n = 152 n = 30 due to falling outside the inclusion criteria; n = 6 currently taking sleep medication; n = 4 difficulty with the English language; n = 48 insufficient time; n = 7 presented sleep disturbances attributable to a medical or psychological problem; Compared to normative sleepers, insomniacs reported longer SOL and WASO and less TST. They also reported lower sleep satisfaction. On the Sleep Quality Index, they had a lower total score (i.e. poorer sleep quality) and scored lower on the questions “how well you slept”, “difficulty falling asleep” and ”early waking and not being able to go back to sleep”. The insomnia group also scored lower on the overall sleep quality rating. Overall, the experimenter administered significantly more prompts when participants were describing a good night than a bad night of sleep. There was no effect of group and no interaction. The meaning of sleep quality among individuals with insomnia and normal sleepers was broadly similar; subjective feelings regarding the day following sleep appeared to be the most important basis for judging sleep quality. All three used procedures implicated tiredness on waking and throughout the day as most consistently associated with sleep quality judgements and two out of the three methods implicated feeling rested, restored, refreshed, replenished on waking and awakenings in the night. A comprehensive assessment of a patient's appraisal of their sleep quality may require an assessment of waking and daytime variables. 72 sleep quality not covered in the existing research literature To compare the insomnia and normal sleeper groups on the meaning of sleep quality. n = 57 not able to be contacted (n = 57). n = 3 completed the first session but did not return to the second session Inclusion in insomnia group: - met criteria for primary insomnia on the Insomnia Diagnostic Interview (IDI) - the problem must been present at least three nights per week for at least one month. Inclusion in normative sleepers group: - not meeting criteria on the IDI - score of ≤7 on the Insomnia Severity Index Characterization of insomnia group: - 18 females, 7 males - 9 participants (17% of the total sample) met criteria for one or more current DSM-IV-TR Axis 1 diagnoses (specific phobia = 4, major depression = 2, generalized anxiety disorder = 1, anorexia = 1, and alcohol abuse = 1). Characterization of normative sleepers group: - 25 females, 7 males - 2 participants (4% of the total sample) met criteria for specific phobia. Procedures (1) “Speak Freely” in which participants described good and poor sleep quality nights; (2) “Sleep Quality Interview” in which participants judged the relative importance of variables included in previous research on sleep quality (applied to 4 normal sleepers and 4 insomnia patients); (3) Sleep quality diary completed over seven consecutive nights. Other measures applied: the Structured Clinical Interview for DSM-IV (SCID); the Insomnia Severity Index (ISI); Beck Depression Inventory (BDI), State The greatest number of people mentioned 5 categories, 4 of which were the same for both groups: - “Motivation to get up or sleep in the morning” - “Tiredness on waking and throughout the day” - “Sleep onset latency” - “Awakenings in the night” And 1 was different: - insomniacs: “Anxiety, worry, and mood on waking and throughout the day” - normative sleepers: “Alertness, clearheadedness, concentration on waking and throughout the day”. Insomniacs were more likely to mention “monitoring” and “body sensations on waking and throughout the day” compared to the normative sleepers and the latter were more likely to mention “memory of sleep” comparatively to insomniacs. For insomniacs, the most important items for judging sleep quality were: - “how well you slept” (Sleep Quality Index item), - “how tired you feel” when you wake up - “how tired you feel” during the day, - “how rested you feel” when you wake up - “how restored you feel” when you wake up. For normative sleepers, the most important items for judging sleep quality were: - “whether you get enough sleep” - “how rested you feel” - “how restored you feel” when you wake up - “how tired you feel” when you wake up - “how alert you feel” throughout the day. 79 Circadian rhythmicity was assessed by the single cosinor method, which was applied to successive 24h runs of data (real-time), starting at midnight real-time. An acrophase was considered only if the cosine curve fitted the data better than a straight line. Only data on phase is used and for days when no significant results were obtained the mean of the adjacent days was substituted. Sleep diary was administered upon awakening and contained 10 items of which most offered 5 response alternatives graded from 1 to 5. “Sleep quality” and “feeling refreshed” were used as global indicators of sleep. Four items formed a sleep quality index, based on their close covariation across time: sleep quality, calmness of sleep, ease of falling asleep and sleep throughout the allotted time. The relationship between subjective and objective sleep parameters was first analyzed through simple intra-individual regression. The resulting correlation coefficients were then averaged across individuals for each par of variables, and the result t-tested against zero correlation, with d.f. = 7 (after z-transformation). To identify the major physiological predictors of good sleep, while taking correlations between predictors into account, multiple regression was used with polysomnograpical variables as predictors and subjective sleep quality as the dependent variable (stepwise multiple regression for key predictors and multiple regression within each individual to test the obtained predictors and the same dependent variable). total sleep time of naps (by 8 minutes) and overestimated nap latency (by 7 minutes). Rated sleep quality increased with increased sleep efficiency and with closeness to the acrophase. The four item sleep quality index showed essentially the same results. The subjective feeling of being refreshed from sleep was also predicted by higher sleep efficiency and less deviation from the acrophase, although less strongly. Rated wase of awakening was predicted by the deviation from the acrophase and by sleep efficiency. The ease increased with closeness to the acrophase but decreased with sleep efficiency. that “rather good” sleep may refer to a sleep efficiency of 87% and above, whereas “rather poor” sleep is applied to an efficiency of 57% or lower. In the same vein, a “rather easy” sleep onset may take 13 minutes or less and a “rather difficult” one may take 43 minutes or more. Early awakenings may be rated “too early” when they leave 60minutes or more in bed before intended rise time. 1989 Psychiatry Research (Buysse et al., 1989) USA The Pittsburgh Sleep Quality Index: A New Instrument for Psychiatric To provide a reliable, valid, and standardized measure of sleep quality. Longitudinal, prospective study. Measure *The PSQI is a self-rated scale composed by 19 questions to be answer by the individual and 5 Analysis of variance (ANOVA) indicated a significant difference in age between groups (F= 5.20, p< O.OOI), with post hoc differences between control subjects and DIMS and DOES patients. - Individuals find PSQI easy to use. - The seven major components of the index, as well as the 19 individual questions, are internally consistent. 80 Practice and Research To discriminate between “good” and “poor” sleepers. To provide an index that is easy for subjects to use and for clinicians and researchers to interpret. To provide a brief, clinically useful assessment of a variety of sleep disturbances that might affect sleep quality. questions rated by the bedpartner or roommate (these latter only used for clinical information). *These I9 items are grouped into 7 component scores, each weighted equally on a 0 to 3 scale. These 7 component scores are summed to yield a global PSQI score, which has a range from 0 to 21; higher scores indicate worse sleep quality. *The 7 components are standardized versions of areas routinely assessed in clinical interviews of patients with sleep/wake complaints, namely: subjective sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbances, use of sleeping medications, and daytime dysfunction. *PSQI items were derived from three sources: - clinical intuition and experience with sleep disorder patients; - a review of previous sleep quality questionnaires reported in the literature; - clinical experience with the instrument during 18 months of field-testing. Study procedures Recruitment from the research studies: - sleep and aging (MH-37869) - nocturnal penile tumescence (MH-40023) - sleep in depression (MH-40023, MH-30915) Study period: 18 months Individuals were excluded if not in a 2-week medication-free interval and if they present some known central nervous system disease. No specific exclusion criteria were used for the clinic sample of sleep-disorder patients. Evaluation for all subjects included: - a complete medical history and physical examination. - a 2-week sleep/wake diary and a sleep habits questionnaire. Male subjects had a lower mean age (46.5 years; SD = 16.7) than female subjects (55.4 years; SD = 18.9) (t = -3.01, p<0.005). Many of the male subjects were involved in studies of nocturnal penile tumescence in depression, while female subjects were participating mainly in studies of sleep, aging, and depression. Age was negatively correlated with the subjective sleep quality (r = -0.22, p < 0.05) and daytime dysfunction (r = -0.29, p < 0.02) component scores in the healthy controls. The PSQI global score and other component scores (sleep latency, sleep duration, habitual sleep efficiency, sleep disturbances, and use of sleeping medications) were not significantly correlated with age. The seven component scores of the PSQI had an overall reliability coefficient (Cronbach’s alpha) of 0.83, indicating a high degree of internal consistency. The largest component-total correlation coefficients were found for habitual sleep efficiency and subjective sleep quality (0.76 for each), and the smallest correlation coefficient was found for sleep disturbances (0.35). The mean component-total correlation coefficient was 0.58. Pearson product-moment correlations between component scores and the PSQI global score were also calculated for the entire group, as well as each group separately. Once again, the strongest - The global scores, component scores, and individual question responses are stable across time. - The validity of the index is supported by its ability to discriminate patients from controls, and, to a more limited degree, by concurrent polysomnographic findings. PSQI was designed to assess clinical samples, while most previous questionnaires have been designed to assess normal sleep habits or entire populations. The PSQI is primarily intended to measure sleep quality and to identify good and bad sleepers, not to provide accurate clinical diagnoses. Nevertheless, responses to specific questions can point the clinician toward areas for further investigation. This is particularly true for the “sleep disturbances” component, which may guide clinical evaluations for specific patients, even though mean scores do not discriminate between groups. Furthermore, a PSQI global score > 5 indicates that a subject is having severe difficulties in at least two areas, or moderate difficulties in more than three areas. The global score is therefore “transparent,” i.e., it conveys information about the severity of the subject’s problem, and the number of problems present, through a single simple measure. PSQI responses were not found to correlate with polysomnographic measures. It is not surprising that subjects differed in subjective and polysomnographic variables, since the PSQl asks for a global 81 - a routine polysomnography. The routine sleep montage included: electroencephalographic (C4, referenced to tied mastoids), electrooculographic (EOG), and electromyographic (submental) leads. Most subjects had additional monitoring for sleep apnea, myoclonus, or nocturnal penile tumescence, dictated by clinical indications or research protocol involvement. All sleep records were scored in 1-min epochs according to standard criteria (Rechtschaffen and Kales, 1968) using Stage 2 sleep onset and standard convention for definition of sleep efficiency (time spent asleep/total recording period). All I48 subjects completed the PSQI on at least one occasion during the course of their clinical and research evaluation. For the majority of subjects (n = 107) the PSQI was completed before sleep studies. For some subjects with stable sleep/wake complaints (n= 41) the PSQI was completed after sleep studies. A subgroup of 91 subjects (43 controls, 22 depressives, and 26 sleep-disorder patients) completed the index a second time (an average of 28.2 days later, range: l-265 days). The second PSQI was completed before any pharmacological treatment began. *Group 1: “good” sleepers n=52, healthy controls without sleep complaints Mean age: 59.9 years (range: 24-83) Male/female ratio: 40/12 *Group 2: “poor” sleepers n=34, patients with major depressive disorder (24 outpatients and 10 inpatients at the Western Psychiatric Institute and Clinic) Mean age: 50.9 years (range: 21-80) Male/female ratio: 25/9 correlations were seen for habitual sleep efficiency and subjective sleep quality. Individual items were also strongly correlated with each other, indicated by a reliability coefficient (Cronbach’s a) of 0.83. Item-total correlation Coefficients ranged from 0.66 for question #9 (enthusiasm to get things done) to 0.20 for item #8 (difficulty staying awake). Pearson product-moment correlations between individual items and the global score ranged from 0.83 (subjective sleep quality) to 0.07 (cough or snore during sleep). 91 patients completed the PSQI on two separate occasions. Paired t-tests for the global PSQI score, as well as the seven individual component scores, showed no significant differences between the two time points. Two differences were noted for depressed patients, who showed a reduction in sleep disturbances and daytime dysfunction. Pearson product-moment correlations again demonstrated stability in global and component scores. The correlation coefficient for global PSQI scores was 0.85 (p <0.001). Component scores had coefficients ranging from 0.84 (sleep latency) to 0.65 (medication use) (p < 0.001 for each component score). Global PSQI scores for each diagnostic group were also significantly correlated between the two testing times, with r’s > 0.40, p < 0.005) for each group. Component scores within each subject group showed more variability across time, but all of these scores were significantly correlated (r’s > 0.35, p < estimate spanning 1 month, and is not sensitive to daily variability. The PSQl’s simplicity and its ability to identify different groups of patients suggest several clinical and research applications in psychiatry and general medical settings. Most fundamentally, it may be used as a simple screening measure to identify cases and controls, or “good” and “poor” sleepers. In a general clinical setting, the PSQI could be used to screen patients for the presence of significant sleep disturbance. In psychiatric settings, the PSQI may identify patients who are likely to have a sleep disturbance concomitant with their psychiatric symptoms. In addition, it may direct the clinician to specific areas of dysfunction that require further investigation. The PSQI could also be used in clinical research and epidemiological studies to identify groups that differ in the quality of their sleep. The PSQI may also have several longitudinal applications in clinical practice and research. 82 All depressed patients met criteria for definite or probable current major depressive disorder. *Group 3: “poor” sleepers n=62, physician-referred outpatients at the Sleep Evaluation Center (SEC) of the Western Psychiatric Institute and Clinic. (Patients were referred to the SEC for assessment of a variety of sleep/wake complaints, but only patients with Disorder of Initiating and Maintaining Sleep (DIMS, n=45) or Disorders of Excessive Somnolence (DOES, n =17) (Association of Sleep Disorders Centers-ASDC, 1979) were included in this study). Mean Age: DIMS_ 44.8 years (range: 20-80); DOES_ 42.2 years (range: 19-57). Male/female ratio: DIMS_ l6/29; DOES_ 8/9 Sleep-disorder patients meeting criteria for DSM-III (American Psychiatric Association, 1980) major depression were excluded from the current study. All depressed patients and healthy controls were assessed with: - Schedule for Affective Disorders and Schizophrenia-Lifetime version (SADS-L). - Research Diagnostic Criteria. - Hamilton Rating Scale for Depression. Sleep-disorder patients were evaluated as described elsewhere (Jacobs et al., 1988) and given preliminary diagnoses according to ASDC nosology. Statistical procedures - Descriptive statistics - ANOVA. - Internal homogeneity (Cronbach’s Alpha, corrected component-total correlation coefficients and Pearson product-moment correlations) - Test-retest reliability (consistency) (paired t-tests and Pearson product-moment correlations for 0.05). The single exception was medication use in control subjects, which showed no correlation between the two testing times. Global PSQI scores differed significantly between subject groups, using an ANCOVA with age and sex as covariates. Control subjects differed from all patient groups (Student-Neuman-Keul’s procedure). Group differences resulted in distinctive component and global score profiles. Age was a significant covariate only for the daytime dysfunction component; but contrary to expectations, these factors were inversely correlated, i.e., reported severity of daytime dysfunction tended to be greater in younger than in older subjects. Sex was a significant covariate for use of sleeping medications and habitual sleep efficiency, with males showing higher scores for each of these components. Age and sex were both significant covariates for the PSQl global score, but group differences were highly statistically significant even after covarying for these factors. A post hoc cutoff score of 5 correctly identified 88.5% (131/148) of all patients and controls (kappa = 0.75, p < 0.001). This represents a sensitivity of 89.6% and a specificity of 86.5%. The same cutoff score correctly identified 84.4% (38/45) of DIMS patients, 88% (IS/ 17) of DOES patients, and 97% (33/34) of depressives. Group differences in PSQI global scores were also substantiated by polysomnographic results, which showed significant group differences for sleep latency (F= 4.53, p < O.OOl), sleep 83 PSQI global score, component scores, and individual items, at Time 1 versus Time 2. - Validity (the degree to which the index detected differences between groups recognized clinically as distinct. This assumes that the index measures differences between groups at the same time point as a clinical “gold standard”. In this case, the relevant “gold standard” diagnoses were based on a combination of clinical interviews, structured interviews, and polysomnographic data. For this analysis, an analysis of covariance (ANCOVA) was used to compare patient groups for PSQI global and component scores, and the Student-NeumanKeul’s procedure was used for pairwise comparisons. Age and sex were used as covariates because of group differences in age and sex ratio. A multiple ANCOVA (MANCOVA) was performed for the PSQI global score, again using age and sex as covariates. A secondary analysis of validity, we compared PSQI scores with polysomnographic results, being cognizant of the fact that PSQI scores reflect the experience of sleep during the previous month, while polysomnographic data were limited to 2 or 3 nights. PSQI estimates of sleep latency, sleep duration, and sleep efficiency were compared to their homologous polysomnographic measures, using both t-tests and Pearson product-moment correlations. Global PSQI scores were also compared to polysomnographic variables selected a priori as being likely to correlate with overall sleep quality, again using Pearson correlations. The specific variables selected were REM Yc, Delta 70, sleep latency, sleep efficiency, and sleep duration. Finally, group differences for these polysomnographic variables were assessed using one-way ANOVAs. efficiency (F = 5.78, p < O.OOl), sleep duration (F= 4.82, p < 0.003), and number of arousals (F= 2.87, p < 0.04). Significant group differences were not found for rapid eye movement (REM) % or delta sleep %. Validity of the PSQI was further examined by comparing PSQI estimates of sleep variables with those obtained by polysomnography. T tests showed no differences between PSQI estimates and laboratory findings for sleep latency, but PSQl estimates of the past month’s usual sleep duration and efficiency were greater than those obtained during polysomnography (t = 9.98 and 4.50, respectively; both p’s < 0.001). This pattern was true for the total subject pool as well as individual subject groups. Pearson correlations demonstrated no significant positive correlations between PSQI estimates and polysomnographic results, except in sleep latency for the total subject pool (r = 0.33, p < 0.001) and for the depressive subgroup (r = 0.37, p < 0.02). Similarly, the global PSQI score was compared with several polysomnographic measures which we selected a priori as being likely to correlate with perceived sleep quality. For all subjects, the global score was weakly correlated only with objective sleep latency (r = 0.20, p < 0.01). For individual subject groups, the global PSQI score correlated only with REM Yc in controls (r = 0.34, p < 0.006) and number of arousals in depressives (r = 0.47, p < 0.002). 84 1987 Res Nurs Health (Snyder‐ Halpern & Verran, 1987) USA Instrumentation to describe subjective sleep characteristics in healthy subjects To develop and test the Verran and SnyderHolpern (VSH) Sleep Scale, an instrument to subjectively measure sleep characteristics. 1971 British Journal of Preventive and Social Medicine (Johns et al., 1971) Australia Sleep Habits of healthy young adults: use of a sleep questionnaire To describe the variations in, and the intercorrelations between, the answers given by students to questions relating to the quality and quantity of their usual sleep. Some of these students took part also in two additional studies on the relationship between levels of adrenocortical activity and sleep habits, and between personality and sleep habits, the results of which will be reported separately. Sample: n=249 in a total of 286 medical students n=122 students in 1969 n=127 in 1970 Cohort characterization: Mean age= 21.5 ± 1.5 yrs; 213 males and 26 females; 230 unmarried, 17 married and 2 divorced. The questionnaire designed asked such questions as: - 'at what time do you usually go to bed at night on weekdays?' - 'how would you describe your usual sleep?' A range of possible answers was provided and the most appropriate were selected by each student. The questionnaire used in 1970 had 27 questions while that in 1969 had 31 questions, all of the important questions being the same. The total delay before falling asleep, duration of night awakenings and of sleep during the night and day, etc., were each calculated in hours per week rather than hours per 24 hours, thereby partially overcoming differences between weekdays and weekends. 85 Discussion In our 24-h society, social pressures weaken and/or suppress biological drives, ultimately, affecting sleep habits and needs. The combination of different social roles, and the high demands inherent to this, often put the individual in the position of sacrificing personal rest and sleep time, increasing the vulnerability to health problems and sleep complaints. The decrease in sleep duration is frequently associated to decreases in sleep quality, a far more heterogeneous and less consensual concept. In fact, when performing a quick search on PubMed®, filtering for one-year results, it is possible to observe not only a high number of articles addressing sleep quality without a definition of what “sleep quality” actually is, but also a high heterogeneity in the assessment tools used. This not only leads to different sleep quality considerations, but consequently to a high variability across studies’ results (Harvey et al., 2008). For some authors sleep quality is a composite measure constituted by different sleep parameters (Amorim et al., 2018; Gupta, Ulfberg, Allen, & Goel, 2018; Kim et al., 2018; Klumpp, Hosseini, & Phan, 2018), as it happens when using the global PSQI score. For others, it concerns one particular parameter, such as the rating of sleep “depth” or of the quality of the sleep of the previous night or month (Gombert, Konze, Rivkin, & Schmidt, 2018; Liu et al., 2018; Takeuchi et al., 2018; Zhang et al., 2018). On this topic, it is also relevant to highlight that these ratings can also be an important source of bias. In fact, across studies, there is a wide range of options concerning the length and type of scales used. For example, while in some studies, visual analog scales are the ones used, in other studies, more qualitative options are considered. This can be within itself quite subjective and bias promoter. In fact, in one of the analyzed studies, individuals had to rate their “depth of sleep” using the following options: “1= can have a sound sleep; 2= can relatively have a sound sleep; 3= neither; 4= relatively bad; 5= very bad” (Takeuchi et al., 2018). However, not only the meaning of “sound sleep” is subjective, but also the value that this has to the participant is of care. In another study, the options used were not mutually exclusive, e.g.: “use of sleeping pills or drugs”, “difficult to fall asleep”, “dreamy sleep”, “can fall asleep but easily awaken”, and “sleep well” (Lao et al., 2018). Nonetheless, quantitative scales also present variability issues important to consider, namely, options ranging from 0 to 5, 0 to 10 or even 0 to 80, depending on the study (Angelhoff et al., 2018; Vitale, Banfi, La Torre, & Bonato, 2018; Zhang et al., 2018). This poses the question of how sure we are that they are all the same, especially because it is unclear how differently individuals interpret these ranges and how reliable they are to translate what the individual think. 86 Furthermore, in some studies an adaptation of existing questionnaires was used (6%) without indication for a proper validation study. For example, in the study of Ko & Lee (2018), a set of questions was considered from Verran and Snyder-Halpern Sleep Scale and in another study (Xie, Dong, & Wang, 2018) the same procedure was done regarding the PSQI, but none of them explored for the validity of this new measure they were considering. Furthermore, when considering the first 100 articles addressing “sleep quality”, it was quite striking the amount of ways sleep quality was being addressed. If in one hand, there was a combination between subjective and objective measures (Montesinos, Castaldo, Cappuccio, & Pecchia, 2018), on the other, there was the use of measures of disturbance, such as insomnia, considered as a measure of sleep quality (or, at least, the lack of it) (Mantua, Helms, Weymann, Capaldi, & Lim, 2018). Thus, with so much variability, to what extent are these studies comparable? Does any of them reflect the actual interpretation that the general population has regarding sleep quality? In a recent report from the National Sleep Foundation (Ohayon et al., 2017), a panel of experts systematically reviewed the literature in order to produce guidelines and recommendations regarding parameters of good sleep quality throughout the lifespan. However, a proper definition of good sleep quality remains elusive. It is still unclear what is the meaning of sleep quality, what constitutes it (i.e. different instruments consider different parameters), in what proportion, and whether it is immutable or not. In fact, sleep quality might be more than just sleep quantity, timing, simple stage structure or occurrence of pathologic events (Krystal & Edinger, 2008). While some authors argue that sleep quality perception often translates one’s satisfaction with his/her sleep (Ohayon et al., 2017), others suggest that daytime functioning is more important for people to judge their own sleep experience when compared to the actual night itself (Goelema et al., 2018). In fact, the only proposed definition of sleep quality is the one from Yi and colleagues (2016) that used (and adapted) the definition from the Oxford English Reference Dictionary – sleep quality is the “degree of excellence in sleep”. However, this is not a satisfying definition, because it still doesn’t answer to the above stated questions and, as Yi and colleagues, stated, no measure can be developed until the nature of the concept has been delineated (Yi et al., 2006). Nevertheless, it should be noted that sleep quality has been measured on the basis of this definition (Freedman, Kotzer, & Schwab, 1999; Hawkins & Shaw, 1992; Shaver, Giblin, & Paulsen, 1991; Yi et al., 2006).To overcome this gap, a qualitative approach is of value . That is, the understanding of how lay people describe and conceptualize sleep holds a relevant role. 87 Here, results indicate that the experienced sleep quality might be more dependent on the daytime functioning than on the progress of the night (attained by retrospective judgment) (Goelema et al., 2018; Harvey et al., 2008; Ramlee et al., 2018). In fact, by conceptualizing sleep quality judgement as a decision-making process, it is possible to quantitatively identify and estimate the relative importance of different sleep and non-sleep parameters influencing this judgement (Ramlee et al., 2018). Among these parameters, participants mostly rely on total sleep time, feeling refreshed upon waking and the mood state in the following day (Ramlee et al., 2017). However, in older studies, results show that subjective sleep quality seems to be essentially associated to perceptions of ease of initiation and maintenance of sleep but unrelated to the perception of the ease of awakening (Keklund & Akerstedt, 1997). In other study of the same group, subjective quality of sleep mainly involved variables of sleep continuity, and "sleep quality", "calm sleep", "ease of falling asleep," and the ability to "sleep throughout'' the time allotted, strongly co-varied and formed an index of sleep quality (Akerstedt et al., 1994). Interestingly, self-rated ease of awakening deviated from the general pattern, as well as reported dreaming (related to awakenings), both associating with poor sleep quality (Akerstedt et al., 1994). Another interesting point, concerns to the variability in the meaning of subjective sleep quality across (clinical) conditions. Not only people with chronic pain conceive pain experience and sleep quality as two linked entities, which influence their ability to engage in planned daytime activities (Ramlee et al., 2018), but also the meaning of sleep quality among individuals with insomnia and normal sleepers, or even between good or poor sleepers, may actually be broadly similar (Harvey et al., 2008; Ramlee et al., 2017). Thus, according to Goelema and colleagues, sleep quality definition, from a subjective point of view, should also consider factors such as stress/well-being levels, rest feeling and functioning during the day, guiding future research in what concerns suitable psychometric measures for normal sleepers, as well as the design of sleep data visualization applications in the context of health selfmonitoring (Goelema et al., 2018). Furthermore, a comprehensive assessment of a patient's appraisal of their sleep quality may require an assessment of waking and daytime variables (Harvey et al., 2008). Importantly, in the study from Keklund & Akerstedt, it is argued that because they created a sufficient variation of sleep quality (including also poor sleep), the needed variance for a correlation between subjective and objective measures was achieved (Keklund & Akerstedt, 1997). In fact, a good agreement between laboratory and field situations with respect to how the subjective items are inter-related, and to the covariation between subjective and objective measures of sleep 88 was observed (Keklund & Akerstedt, 1997). Considering the association between objective sleep parameters in subjective sleep quality, it has been shown that subjective sleep quality was related mainly to sleep efficiency (accounted for most of the variance) but also to the closeness of the awakening to the circadian acrophase (Åkerstedt et al., 1994). In this particular study, sleep efficiency was more central to “good sleep” than continuity variables, like sleep latency, final wake time and time awake within sleep, which makes sense given that sleep efficiency represents the combination of these variables (Åkerstedt et al., 1994). Furthermore, sleep architecture (e.g. SWS and REM) apparently lacked relevance in subjective sleep quality, at least in what concerns sleep length (Åkerstedt et al., 1994). Interestingly, maximum sleep quality was rated when sleep ended close to the acrophase, which was suggested to be due to the circadian variation of alertness that reaches an acrophase in the early evening (Åkerstedt et al., 1994). The indication is that in future studies, circadian phase should be considered in the interpretation of sleep quality ratings (Åkerstedt et al., 1994). The limitations associated to the small samples and restriction to mostly college students are obvious, as they leave out the population that frequently complaints about their sleep. What does a good sleep quality mean for the general population? What words do they use? Is it dependent on cultural issues? What is the association between this definitions and standard measures of sleep quality, such as the PSQI? What are the determinants of a subjective good sleep quality? These are still unanswered questions. Future studies should address for this in community-dwellers, across the adult lifespan, so to develop a new measure that can be used as standard, not only for research but also for clinicians, permitting for further associations with standard measures, such as the Pittsburgh Sleep Quality Index (PSQI), should also be determined. Overall, sleep quality can be defined as a multidimensional concept constituted by a set of different dimensions, such as duration, efficiency and continuity. Despite using these different dimensions as proxies to good or poor sleep, it is not adequate to refer to any of these as “sleep quality” if only one dimension is being considered. The weight of each parameter in the overall measure of sleep quality is yet to be determined. Thus, studies must start to address the importance of each parameter in not only determining a final composite measure, but also in identifying for possible individual profiles (an adjusted measured). 95 Kim, M.-J., Yim, G., & Park, H.-Y. (2018). Vasomotor and physical menopausal symptoms are associated with sleep quality. PLoS ONE , 13 (2). https://doi.org/10.1371/journal.pone.0192934 Kim, S. H., Kim, H., Kim, S., Paek, S., Koh, J. H., Lee, J., … Park, S.-H. (2018). Sleep Quality Independently Affects Health-related Quality of Life and Cognitive Function in Korean Female Patients with Rheumatoid Arthritis: A Case-control Study. Journal of Korean Medical Science , 33 (35). https://doi.org/10.3346/jkms.2018.33.e216 Kim, S. K., Kim, J. H., Jeon, S. S., & Hong, S. M. (2018). Relationship between sleep quality and dizziness. PLOS ONE , 13 (3), e0192705. https://doi.org/10.1371/journal.pone.0192705 Klumpp, H., Hosseini, B., & Phan, K. L. (2018). Self-Reported Sleep Quality Modulates Amygdala Resting-State Functional Connectivity in Anxiety and Depression. Frontiers in Psychiatry , 9 . https://doi.org/10.3389/fpsyt.2018.00220 Ko, Y., & Lee, J.-Y. (2018). Effects of feet warming using bed socks on sleep quality and thermoregulatory responses in a cool environment. Journal of Physiological Anthropology , 37 . https://doi.org/10.1186/s40101-018-0172-z Kong, F., Li, H., Xu, G., Ying, Y., Gong, Q., Zhao, J., … Han, L. (2018). Association of Dietary Behaviors and Sleep Quality: Results from the Adults Chronic Diseases and Risk Factors Survey of 2015 in Ningbo, China. International Journal of Environmental Research and Public Health , 15 (9). https://doi.org/10.3390/ijerph15091823 Kotterba, S., Neusser, T., Norenberg, C., Bussfeld, P., Glaser, T., Dörner, M., & Schürks, M. (2018). Sleep quality, daytime sleepiness, fatigue, and quality of life in patients with multiple sclerosis treated with interferon beta-1b: Results from a prospective observational cohort study. BMC Neurology , 18 . https://doi.org/10.1186/s12883-018-1113-5 Krističević, T., Štefan, L., & Sporiš, G. (2018). The Associations between Sleep Duration and Sleep Quality with Body-Mass Index in a Large Sample of Young Adults. International Journal of Environmental Research and Public Health , 15 (4). https://doi.org/10.3390/ijerph15040758 Krystal, A. D., & Edinger, J. D. (2008). Measuring sleep quality. Sleep Medicine , 9 , S10–S17. https://doi.org/10.1016/S1389-9457(08)70011-X 96 Kwok, C. S., Kontopantelis, E., Kuligowski, G., Gray, M., Muhyaldeen, A., Gale, C. P., … Mamas, M. A. (2018). Self‐Reported Sleep Duration and Quality and Cardiovascular Disease and Mortality: A Dose‐Response Meta‐Analysis. Journal of the American Heart Association: Cardiovascular and Cerebrovascular Disease , 7 (15). https://doi.org/10.1161/JAHA.118.008552 Kwon, Y. (2018). Atrial Fibrillation and Objective Sleep Quality by Slow Wave Sleep. Journal of Atrial Fibrillation , 11 (2). https://doi.org/10.4022/jafib.2031 Lao, X. Q., Liu, X., Deng, H.-B., Chan, T.-C., Ho, K. F., Wang, F., … Yeoh, E.-K. (2018). Sleep Quality, Sleep Duration, and the Risk of Coronary Heart Disease: A Prospective Cohort Study With 60,586 Adults. Journal of Clinical Sleep Medicine : JCSM : Official Publication of the American Academy of Sleep Medicine , 14 (1), 109–117. https://doi.org/10.5664/jcsm.6894 Lee, A. J., Kraemer, D. F., Kanar, O., Berry, A. C., Smotherman, C., & Eid, E. (2018). Immunomodulator and Biologic Agent Effects on Sleep Quality in Patients With Inflammatory Bowel Disease. The Ochsner Journal , 18 (1), 76–80. Li, C., Kee, Y. H., & Lam, L. S. (2018). Effect of Brief Mindfulness Induction on University Athletes’ Sleep Quality Following Night Training. Frontiers in Psychology , 9 . https://doi.org/10.3389/fpsyg.2018.00508 Li, P., Huang, Y., Guo, L., Wang, W., Xi, C., Lei, Y., … Lu, C. (2017). Is sexual minority status associated with poor sleep quality among adolescents? Analysis of a national crosssectional survey in Chinese adolescents. BMJ Open , 7 (12). https://doi.org/10.1136/bmjopen-2017-017067 Lim, J., & Dinges, D. F. (2010). A meta-analysis of the impact of short-term sleep deprivation on cognitive variables. Psychological Bulletin , 136 (3), 375–389. https://doi.org/10.1037/a0018883 Liu, X., Xiao, S., Zhou, L., Hu, M., Zhou, W., & Liu, H. (2018). Sleep quality and covariates as predictors of pain intensity among the general population in rural China. Journal of Pain Research , 11 , 857–866. https://doi.org/10.2147/JPR.S156731 Mantua, J., Helms, S. M., Weymann, K. B., Capaldi, V. F., & Lim, M. M. (2018). Sleep Quality and Emotion Regulation Interact to Predict Anxiety in Veterans with PTSD. Behavioural Neurology , 2018 . https://doi.org/10.1155/2018/7940832 97 Manzar, M. D., BaHammam, A. S., Hameed, U. A., Spence, D. W., Pandi-Perumal, S. R., Moscovitch, A., & Streiner, D. L. (2018). Dimensionality of the Pittsburgh Sleep Quality Index: A systematic review. Health and Quality of Life Outcomes , 16 . https://doi.org/10.1186/s12955-018-0915-x Michalopoulos, G., Vrakas, S., Makris, K., & Tzathas, C. (2018). Association of sleep quality and mucosal healing in patients with inflammatory bowel disease in clinical remission. Annals of Gastroenterology , 31 (2), 211–216. https://doi.org/10.20524/aog.2018.0227 Montesinos, L., Castaldo, R., Cappuccio, F. P., & Pecchia, L. (2018). Day-to-day variations in sleep quality affect standing balance in healthy adults. Scientific Reports , 8 (1), 17504. https://doi.org/10.1038/s41598-018-36053-4 Moudi, A., Dashtgard, A., Salehiniya, H., Sadat Katebi, M., Reza Razmara, M., & Reza Jani, M. (2018). The relationship between health-promoting lifestyle and sleep quality in postmenopausal women. BioMedicine , 8 (2). https://doi.org/10.1051/bmdcn/2018080211 Moul, D. E., Nofzinger, E. A., Pilkonis, P. A., Houck, P. R., Miewald, J. M., & Buysse, D. J. (2002). Symptom reports in severe chronic insomnia. Sleep , 25 (5), 553–563. Musa, N. A., Moy, F. M., & Wong, L. P. (2018). Prevalence and factors associated with poor sleep quality among secondary school teachers in a developing country. Industrial Health , 56 (5), 407–418. https://doi.org/10.2486/indhealth.2018-0052 Ohayon, M., Wickwire, E. M., Hirshkowitz, M., Albert, S. M., Avidan, A., Daly, F. J., … Vitiello, M. V. (2017). National Sleep Foundation’s sleep quality recommendations: First report. Sleep Health: Journal of the National Sleep Foundation , 3 (1), 6–19. https://doi.org/10.1016/j.sleh.2016.11.006 Raikes, A. C., Bajaj, S., Dailey, N. S., Smith, R. S., Alkozei, A., Satterfield, B. C., & Killgore, W. D. S. (2018). Diffusion Tensor Imaging (DTI) Correlates of Self-Reported Sleep Quality and Depression Following Mild Traumatic Brain Injury. Frontiers in Neurology , 9 . https://doi.org/10.3389/fneur.2018.00468 Ramlee, F., Afolalu, E. F., & Tang, N. K. Y. (2018). Do People With Chronic Pain Judge Their Sleep Differently? A Qualitative Study. Behavioral Sleep Medicine , 16 (3), 259–271. https://doi.org/10.1080/15402002.2016.1188393 98 Ramlee, F., Sanborn, A. N., & Tang, N. K. Y. (2017). What Sways People’s Judgment of Sleep Quality? A Quantitative Choice-Making Study With Good and Poor Sleepers. Sleep , 40 (7). https://doi.org/10.1093/sleep/zsx091 Rebello, C. R., Kallingappa, P. B., & Hegde, P. G. (2018). Assessment of perceived stress and association with sleep quality and attributed stressors among 1st-year medical students: A cross-sectional study from Karwar, Karnataka, India. Tzu-Chi Medical Journal , 30 (4), 221–226. https://doi.org/10.4103/tcmj.tcmj_85_17 Risom, S. S., Fevejle Cromhout, P., Overgaard, D., Hastrup Svendsen, J., & Kikkenborg Berg, S. (2018). Effect of Rehabilitation on Sleep Quality After Ablation for Atrial Fibrillation. The Journal of Cardiovascular Nursing , 33 (3), 261–268. https://doi.org/10.1097/JCN.0000000000000476 Rogers, A., Morrison, I., Rorie, D. A., Mackenzie, I. S., & MacDonald, T. M. (2018). Protocol for assessment of sleep quality and duration in the Treatment In Morning versus Evening (TIME) study: A randomised controlled trial using online patient-reported outcome measures. BMJ Open , 8 (6). https://doi.org/10.1136/bmjopen-2018-021890 Rönnlund, M., & Carelli, M. G. (2018). Time Perspective Biases Are Associated With Poor Sleep Quality, Daytime Sleepiness, and Lower Levels of Subjective Well-Being Among Older Adults. Frontiers in Psychology , 9 . https://doi.org/10.3389/fpsyg.2018.01356 Rosa, R. R., & Bonnet, M. H. (2000). Reported Chronic Insomnia Is Independent of Poor Sleep as Measured by Electroencephalography. Psychosomatic Medicine , 62 (4), 474. Rosipal, R., Lewandowski, A., & Dorffner, G. (2013). In search of objective components for sleep quality indexing in normal sleep. Biological Psychology , 94 (1), 210–220. https://doi.org/10.1016/j.biopsycho.2013.05.014 Sakamoto, R., Yamakawa, T., Takahashi, K., Suzuki, J., Shinoda, M. M., Sakamaki, K., … Terauchi, Y. (2018). Association of usual sleep quality and glycemic control in type 2 diabetes in Japanese: A cross sectional study. Sleep and Food Registry in Kanagawa (SOREKA). PLoS ONE , 13 (1). https://doi.org/10.1371/journal.pone.0191771 Shaver, J. L. F., Giblin, E., & Paulsen, V. (1991). Sleep Quality Subtypes in Midlife Women. Sleep , 14 (1), 18–23. https://doi.org/10.1093/sleep/14.1.18 Simon, E. B., & Walker, M. P. (2018). Sleep loss causes social withdrawal and loneliness. Nature Communications , 9 (1), 3146. https://doi.org/10.1038/s41467-018-05377-0 99 Simons, K. S., Verweij, E., Lemmens, P. M. C., Jelfs, S., Park, M., Spronk, P. E., … de Jager, C. P. C. (2018). Noise in the intensive care unit and its influence on sleep quality: A multicenter observational study in Dutch intensive care units. Critical Care , 22 . https://doi.org/10.1186/s13054-018-2182-y Simor, P., Harsányi, A., Csigó, K., Miklós, G., Lázár, A. S., & Demeter, G. (sem data). Eveningness is associated with poor sleep quality and negative affect in obsessive–compulsive disorder. Journal of Behavioral Addictions , 7 (1), 10–20. https://doi.org/10.1556/2006.7.2018.07 Snyder‐Halpern, R., & Verran, J. A. (1987). Instrumentation to describe subjective sleep characteristics in healthy subjects. Research in Nursing & Health , 10 (3), 155–163. https://doi.org/10.1002/nur.4770100307 Song, T.-J., Cho, S.-J., Kim, W.-J., Yang, K. I., Yun, C.-H., & Chu, M. K. (2018). Poor sleep quality in migraine and probable migraine: A population study. The Journal of Headache and Pain , 19 (1). https://doi.org/10.1186/s10194-018-0887-6 Štefan, L., Sporiš, G., Krističević, T., & Knjaz, D. (2018). Associations between sleep quality and its domains and insufficient physical activity in a large sample of Croatian young adults: A cross-sectional study. BMJ Open , 8 (7). https://doi.org/10.1136/bmjopen-2018-021902 Stephan, Y., Sutin, A. R., Bayard, S., Križan, Z., & Terracciano, A. (2018). Personality and Sleep Quality: Evidence from Four Prospective Studies. Health psychology : official journal of the Division of Health Psychology, American Psychological Association , 37 (3), 271–281. https://doi.org/10.1037/hea0000577 Sun, T., Gao, L., Li, F., Shi, Y., Xie, F., Wang, J., … Fan, L. (2017). Workplace violence, psychological stress, sleep quality and subjective health in Chinese doctors: A large crosssectional study. BMJ Open , 7 (12). https://doi.org/10.1136/bmjopen-2017-017182 Takeuchi, H., Taki, Y., Nouchi, R., Yokoyama, R., Kotozaki, Y., Nakagawa, S., … Kawashima, R. (2018). Shorter sleep duration and better sleep quality are associated with greater tissue density in the brain. Scientific Reports , 8 (1), 5833. https://doi.org/10.1038/s41598018-24226-0 Teker, A. G., & Luleci, N. E. (2018). Sleep quality and anxiety level in employees. Northern Clinics of Istanbul , 5 (1), 31–36. https://doi.org/10.14744/nci.2017.58855 100 Thichumpa, W., Howteerakul, N., Suwannapong, N., & Tantrakul, V. (2018). Sleep quality and associated factors among the elderly living in rural Chiang Rai, northern Thailand. Epidemiology and Health , 40 . https://doi.org/10.4178/epih.e2018018 Tse, A. C. Y., Lee, P. H., Zhang, J., & Lai, E. W. H. (2018). Study protocol for a randomised controlled trial examining the association between physical activity and sleep quality in children with autism spectrum disorder based on the melatonin-mediated mechanism model. BMJ Open , 8 (4). https://doi.org/10.1136/bmjopen-2017-020944 Tsou, Y.-A., Huang, C.-W., Wu, T.-F., Hung, L.-W., & Chang, W.-D. (2018). The effect of tongue base suspension with uvulopalato-pharyngoplasty on sleep quality in obstructive sleep apnea. Scientific Reports , 8 . https://doi.org/10.1038/s41598-018-27094-w Tuyan İlhan, T., Uçar, M. G., Gül, A., Saymaz İlhan, T., Yavaş, G., & Çelik, Ç. (2017). Sleep quality of endometrial cancer survivors and the effect of treatments. Turkish Journal of Obstetrics and Gynecology , 14 (4), 243–248. https://doi.org/10.4274/tjod.59265 van Eerde, W., & Venus, M. (2018). A Daily Diary Study on Sleep Quality and Procrastination at Work: The Moderating Role of Trait Self-Control. Frontiers in Psychology , 9 , 2029. https://doi.org/10.3389/fpsyg.2018.02029 Vitale, J. A., Banfi, G., La Torre, A., & Bonato, M. (2018). Effect of a Habitual Late-Evening Physical Task on Sleep Quality in Neither-Type Soccer Players. Frontiers in Physiology , 9 . https://doi.org/10.3389/fphys.2018.01582 Wu, W., Wang, W., Dong, Z., Xie, Y., Gu, Y., Zhang, Y., … Tan, X. (2018). Sleep Quality and Its Associated Factors among Low-Income Adults in a Rural Area of China: A Population-Based Study. International Journal of Environmental Research and Public Health , 15 (9). https://doi.org/10.3390/ijerph15092055 Xie, L., Kang, H., Xu, Q., Chen, M. J., Liao, Y., Thiyagarajan, M., … Nedergaard, M. (2013). Sleep Drives Metabolite Clearance from the Adult Brain. Science (New York, N.Y.) , 342 (6156). https://doi.org/10.1126/science.1241224 Xie, X., Dong, Y., & Wang, J. (2018). Sleep quality as a mediator of problematic smartphone use and clinical health symptoms. Journal of Behavioral Addictions , 7 (2), 466–472. https://doi.org/10.1556/2006.7.2018.40 Yi, H., Shin, K., & Shin, C. (2006). Development of the Sleep Quality Scale. Journal of Sleep Research , 15 (3), 309–316. https://doi.org/10.1111/j.1365-2869.2006.00544.x 101 Zhang, S.-E., Liu, W., Wang, J., Shi, Y., Xie, F., Cang, S., … Fan, L. (2018). Impact of workplace violence and compassionate behaviour in hospitals on stress, sleep quality and subjective health status among Chinese nurses: A cross-sectional survey. BMJ Open , 8 (10). https://doi.org/10.1136/bmjopen-2017-019373 Zou, L., Yeung, A., Quan, X., Boyden, S. D., & Wang, H. (2018). A Systematic Review and MetaAnalysis of Mindfulness-Based (Baduanjin) Exercise for Alleviating Musculoskeletal Pain and Improving Sleep Quality in People with Chronic Diseases. International Journal of Environmental Research and Public Health , 15 (2). https://doi.org/10.3390/ijerph15020206 Zwart, T. C., Smits, M. G., Egberts, T. C. G., Rademaker, C. M. A., & van Geijlswijk, I. M. (2018). Long-Term Melatonin Therapy for Adolescents and Young Adults with Chronic Sleep Onset Insomnia and Late Melatonin Onset: Evaluation of Sleep Quality, Chronotype, and Lifestyle Factors Compared to Age-Related Randomly Selected Population Cohorts. Healthcare , 6 (1). https://doi.org/10.3390/healthcare6010023 102 CHAPTER III Sleep quality meaning and its association with the Pittsburgh Sleep Quality Index (PSQI): a population-based study across the adult lifespan Liliana Amorim, Nuno Sousa, Nadine Correia Santos (In preparation to be submitted) 103 Sleep quality meaning and its association with the Pittsburgh Sleep Quality Index (PSQI): a population-based study across the adult lifespan Authors: Liliana Amorim1,2,3, Nuno Sousa1,2,3, Nadine Correia Santos1,2,3 1Life and Health Sciences Research Institute (ICVS), School of Medicine, University of Minho, Campus Gualtar, Braga, Portugal. 2ICVS/3B’s - PT Government Associate Laboratory, Braga/Guimarães, Portugal. 3Clinical Academic Center – Braga, Braga, Portugal. *Corresponding author: Nadine Correia Santos, Life and Health Sciences Research Institute (ICVS), School of Medicine, University of Minho, Campus Gualtar, 4710-057 Braga, Portugal. Email: [email protected]. Phone: +351 253 604 806. 104 Abstract Sleep is a complex, dynamic and multidimensional phenomenon, whose impact has been demonstrated in health and well-being. While the dimension ‘sleep quantity’ is clearly defined, such does not apply to the ‘sleep quality’ construct. In fact, different conceptualizations of it have led to a high variability in the methodologies used and, consequently, in the results obtained across studies. Likewise, authors still have not addressed the meaning of a good sleep quality across the adult lifespan, nor its stability over time. Here, to addresses these aspects, and explore the association of sleep quality self-ratings with the Pittsburgh Sleep Quality Index (PSQI), communitydwellers were invited to participate in the study (n=343, after inclusion and exclusion criteria; age range: 18-87 years). Participants were asked to provide their interpretation of a good sleep quality and a good night of sleep. Self-ratings of sleep quality, patterns and habits and information on psychological variables were also collected. Content analysis was performed with the qualitative data and the obtained parameters were considered as to their frequency. Associations between self-rated sleep quality and PSQI were determined, and the predictors of good/poor sleep quality explored. Results show that the distribution of the reported parameters used to describe a ‘good night of sleep‘ and a ‘good sleep quality‘ differs. While the most frequently reported parameters for ‘good sleep quality’ are ‘sleep continuity’, ‘sleep characteristics’ and a dyad involving these two, for a ‘good night of sleep’ the most reported information related to the dyad ‘sleep duration-sleep continuity’ and to ‘sleep duration’ and ‘sleep continuity’ individually. Results also indicate that selfrated sleep quality measures and the PSQI are highly correlated, and no differences were found across lifespan in the distribution of these parameters. One year later, 52% of the invited individuals who were re-evaluated maintained the definition provided in the first assessment moment. The results contribute to the clarification of the meaning of ‘a good sleep quality’, and provide a relevant framework for future studies addressing the complaints and sleep changes across lifespan. Keywords: Sleep quality; definition; determinants; content analysis; lifespan; network analysis. 111 values (e.g. regularized partial correlations) of a given sleep parameter with all directly related sleep parameters (i.e. the sum of the absolute values of the sleep interrelations). The expected influence is based on the formula of node strength, but takes negative relationships between sleep parameters into account (i.e. the sum of the relative values of the sleep interrelations). Node predictability is defined as the amount of variance of each sleep parameters that is explained by the directly related sleep parameters. Node predictability is an absolute metric ranging from zero to 100 percent explained variance. For dichotomous sleep parameters, we based the node predictability on the normalized accuracy, instead of on the variance explained. Accuracy can be scrutinized through calculating nonparametric bootstrap confidence intervals (CIs, 95%) for the sleep parameters interrelations. The widths of these CIs give an indication for accuracy. Stability can be analyzed through re-calculating interrelatedness coefficients such as the node strength for sample subsets. If the node strength remains similar in the subsets, this indicates that the sleep parameters network is stable. Accordingly, we bootstrapped the sleep parameters interrelations (i.e. accuracy) and applied a subset bootstrap on node strength and expected influence (i.e. stability), with 300 bootstraps each. As sensitivity analysis, we correlated the sleep parameters interrelations of the full-information networks with the sleep parameters interrelations of the complete-information networks, which are based on listwise case deletion. The stability of the parameters was determined by calculating the percentage of individuals that used different parameters to report the meaning of good sleep quality. The association between the different measurements of sleep quality was addressed using the Spearman’s correlation test. In order to test the differences between individuals in good and poor sleep quality groups, qualitative and quantitative and PSQI measures were used to form the groups. The first step was to convert all scales in order for them to be in the same direction (PSQI – higher scores, lower sleep quality, self-reports used – higher scores, better sleep quality). For this, qualitative and quantitative self-reports were converted in a direct manner, i.e., scales were inverted. Then, for the qualitative self-report, the options “really bad” and “more or less bad” were converted in 1 and “very good” and “more or less good” were converted in 0. For the qualitative scale (range: 0 to 100), direct inversion of the scale first and then the ROC method was applied (PSQI as Standard Measure) and 27.5 was the cut-off value for good sleep quality (see Appendix 1 for more details). Thus, all values bellow 27.5 were considered 0 (good sleep quality) and all values above were 112 defined as 1 (poor sleep quality). For PSQI, it was considered the validated cut-off of 5 (Del Rio João et al., 2017) and the PSQI global score was considered in its dichotomous form. Differences between individuals with good and poor sleep quality were then tested for each method of group derivation – PSQI global score, qualitative self-report and quantitative self-report - using ManWhitney test. After determining the differences between groups, the variables of interest were used as determinants of good sleep quality in logistic regression models. All tests were corrected for multiple comparisons. Results Socio demographic and psychological characterization of the cohort A total of n=343 individuals agreed to participate in the study (63.7% female; median age 47 years (IQR=23); 66% married; 38% thirteen or more years of education; 67% employed). However, because of time constraints, only n=308 individuals constituted the qualitative study sample (the sample did not significantly differ from the original sample in the socio-demographic characteristics listed). From the participants able to provide information regarding their psychological status, the median value of subjective sleep quality is 6 (IQR=5) (all values above 5 indicates poor sleep quality according to validation studies) (Buysse et al., 1989; Del Rio João et al., 2017), and 46.3% had poor sleep quality. The median value of sleepiness was 7 (IQR=5) (equal or above 10 it is considered excessive sonmnolence), and stress was the dimension of psychological morbidity with the highest median (median=5; IQR=5). Socio-demographic characteristics, subjective sleep, personality and psychological parameters characterization are shown in Table 1. Table 1. Sociodemographic and psychological characterization of the cohort. n (%) Median (IQR) Age -- 47 (23) ¥ Sex Female 216 (63.7%) Male 127 (36.3%) Education No education 4 (1.2%) 0 to 4 yrs 39 (11.6%) 5 to 9 yrs 74 (22%) 10 to 12 yrs 92 (27.4%) 13 or more yrs 127 (37.8%) Marital Status Single 85 (24.8%) Married 228 (66.5%) Divorced 22 (6.4%) Widowed 8 (2.3%) Household Alone 20 (5.9%) Husband/Wife 133 (39.1%) Family# 183 (53.8%) Others* 4 (1.2%) 113 Occupation Retired 48 (14.1%) Employed 230 (67.4%) Unemployed 29 (8.5%) Student 28 (8.2%) House keeper 6 (1.8%) Living Place City 193 (56.6%) Rural 148 (43.4%) Children in School Yes 134 (39.9%) Grandchildren in School Yes 47 (14.3%) PSQI£ Good Poor 105 (46.3%) 122 (53.7%) 4 (2) ¥ 9 (4) ¥ ESS 118 (--) 7.55 (±3.94) DASS-21 Stress Anxiety Depression 172 (--) 5 (5) ¥ 1 (3.25) ¥ 1 (4) ¥ PANAS Positive Affect Negative Affect 112 (--) 33.3 (6.13) 17 (8) ¥ NEO-FFI Neuroticism Extroversion Openness to experience Agreeableness Consciousness 116 (--) 6.05 (2.78) 9.91 (2.55) 9.95 (3.33) 10-6 (2.78) 12 (2) ¥ ¥Normality and symmetry assumptions not filled, thus, values presented are for median and IQR. #Family - includes any family members, e.g. children, parents, grandparents; *Others – e.g. friends, boyfriend/girlfriend, professional residencies. PSQI – Pittsburgh Sleep Quality Index (£PSQITotal population: median= 6; IQR=5); ESS – Epworth Sleepiness Scale; DASS-21 – Depression, Anxiety and Stress Scale 21 items; PANAS – Positive and Negative Affect Scale; NEO-FFI-21 – 5 Factors Personality Scale Sleep patterns and bedtime routines characterization throughout adult lifespan Regarding sleep schedules, results show that participants go to bed at 23:00 (median) (IQR=1), turn off the lights around the same time (IQR=1), take 15 minutes to fall asleep (IQR=25), wake up at 7:00 (IQR=3), rise 10 minutes (IQR=17) later and report having slept 7 hours (IQR=2). In terms of patterns of routines when going to bed, of the n=326 participants that provided information regarding these parameters, 62.3% indicate having some type of activity in bed, with the majority reporting the use of the mobile phone (Figure 1). The median time spent in activities after going to bed was of 40 minutes (IQR=17). In what concerns the use of alarm clock to wake up, 46.3% of the participants reported its use in every day of the week and at weekends, and 34.9% indicate it would depend on the circumstances and of the day (week or weekend days). No differences were found between men and women in what concerns bedtime routines (χ2(7) =7.88, p=0.344). When exploring the effect of age, a small effect was observed on subjective sleep quality (rho=0.188, p=0.005) and on the reported amount of sleep (rho=-0.297, p=0.001), with no significant impact on sleep latency (rho=0.079, p=0.189). A moderate effect of age was observed regarding sleep (rho=-0.331, p<0.001) and wake times (rho=-0.512, p<0.001) (Figure 2). 114 Figure 1. Bedtime routines across lifespan. 0 0 20 22 24 26 28 30 20 30 40 50 60 70 80 90 Age Bedtime (Time) 0 0 20 22 24 26 28 30 20 30 40 50 60 70 80 90 Age Time Lights Off Bed Time 0 0 20 40 60 80 100 120 140 160 180 20 30 40 50 60 70 80 90 Age Sleep Latency (minutes) 0 0 2 4 6 8 10 12 14 20 30 40 50 60 70 80 90 Age Wake Up Time (Time) 0 0 2 4 6 8 10 12 20 30 40 50 60 70 80 90 Age Sleep Duration (hours) 0 0 25 50 75 100 125 150 175 200 20 30 40 50 60 70 80 90 Age Time spent in bedtime routines (minutes) Figure 2. Sleep patterns across lifespan. 115 A good sleep quality for a lay individual – content analysis results The content analysis resulted in seven units of analysis/categories: “sleep duration”, “sleep continuity”, “sleep onset”, “sleep characteristics”, “room environment”, “final wake up” and “next day performance” (see Table 2 for more details). It was determined, for each unit of analysis, the percentage of individuals that referred it. The same was performed regarding dyads of units and units grouped in one response. A graphic representation of these results is presented in Figures 3 and 4. Answers varied according to the question being about a good night of sleep or a good sleep quality. Particularly, when asked about their night of sleep, participants more frequently reported the dyad “sleep duration”-“sleep continuity” and individually, the units “sleep duration” and “sleep continuity” (Figure 3). The distribution of the answers changed for the question related to a good sleep quality, where inquirers most frequently reported “sleep continuity” aspects alone (Figure 4). It was also common to observe answers related with the individual unit “sleep characteristics” and the dyad “sleep continuity”-“sleep characteristics” (Figure 4). Table 2. Operationalization of the units of analyze resultant of the 604 statements regarding the meaning of “a good night of sleep” and “a good sleep quality”. Units of analyses Description Examples Sleep duration Contains all the information referring to the amount of time the subject sleep. “Sleeping at least 8h”; “Sleeping all night”; “Sleep the appropriate number of hours”. Sleep continuity Contains information about the continuity of sleep and interruptions throughout the night. Sleep fragmentation can be either by internal causes (e.g. need to go to the toilet) and external causes (e.g. children needing assistance or noise). “Sleep in a continuous way”; “Sleep 8-h without interruptions”; “Sleep without dreams/nightmares interruptions and awaking”; “Only wake up once through the night to go to the bathroom”. Sleep onset Contains information regarding sleep latency (i.e. the amount of time to fall asleep) and statements regarding factors that can relate to the amount of time to fall asleep, like pain, worries or the use of medication. “Fall asleep without difficulties”; “Fall asleep quickly”; “Fall asleep peacefully”; “Fall asleep without bad thoughts or worries”; “Fall asleep without pills”; “Go to bed and immediately fall asleep”. Sleep characteristics Contains the subjective perceptions regarding sleep as well as mentions to dreams and nightmares. “To be able to relax physically and mentally”; “A peaceful sleep”; “Sleep well”; “Have a refreshing and peaceful sleep”; “To be able to rest body and mind”; “Restful sleep”; “Sleep with dreams”; “Sleep without dreams or nightmares”. Room environment Contains statements about the physical elements in the bedroom, like the mattress, pillow, lights and temperature but also aspects like noise and the comfortableness of the individual in the bed. “Quiet room”; “Being in a comfortable position”; “Without noises, with a temperature of approximately 19ºC and a dark bedroom, so that it is possible to sleep peacefully”; “Comfortable”; “In silence”. Final wake up and rise Contains statements that inform about the waking up. Specifically, how the participants feel waking up and whether the waking is with or without an alarm clock. “Wake up with rest body and mind”; “Wake up in the morning with felling of being well”; “Wake up naturally, without alarm clock”; Wake up with the alarm clock”; “Wake up with energy”; “Wake up refreshed”; “Wake up in a good mood”. Next day performance Contains information about daytime functioning and statements that mention broad consequences of a good sleep. “Enjoy the day in its fullness”; “Be ready for the day without feeling tired”; “Having quality of life”; “Not being sleepy throughout the day”; “Being calm throughout the day”; “Being productive during the day”; “Being in a good health”; “Don’t feel tired through the day. 116 Figure 3. Graphical representation of the reported parameters for a good night of sleep. 1A is a schematic representation of parameters reported in the global analysis for the question: “For you, what is a good night of sleep?”. Each node represents one unit of analysis and its size reflects the number of participants that have reported it. The connection between nodes represent the dyads of parameters that were simultaneously reported and its with reflects the amount of participants referring it. In 1B, it is represented the number of individuals that report information in each categories and the dyads of observed units of analysis. This table is the raw data considered to develop 1A. In 1C, it is observable the number of individuals that report more than two units of analysis when answering to the question. FIGURE LEGEND 1. Sleep duration 2. Sleep continuity 3. Sleep characteristics 4. Sleep onset 5. Room environment 6. Final wake up and rise 7. Next day performance 6 3 5 1 2 7 4 A 117 Figure 4. Graphical representation of the reported parameters for a good sleep quality. 2A is a schematic representation of parameters reported in the global analysis for the question: “For you, what is a good sleep quality?”. Each node represents one unit of analysis and its size reflects the number of participants that have reported it. The connection between nodes represent the dyads of parameters that were simultaneously reported and its with reflects the amount of participants referring it. In 2B, it is represented the number of individuals that report information in each categories and the dyads of observed units of analysis. This table is the raw data considered to develop 1A. In 2C, it is observable the number of individuals that report more than two units of analysis when answering to the question. FIGURE LEGEND 1. Sleep duration 2. Sleep continuity 3. Sleep characteristics 4. Sleep onset 5. Room environment 6. Final wake up and rise 7. Next day performance 1 6 5 2 3 7 4 A 118 Network analysis on statistical structure of the answers from content analysis – a descriptive approach Network analysis (Fried et al., 2015; Schmittmann et al., 2013) was used to obtain relations between the variables and information about their clustering (Figure 5). Network Plot Centrality Plot Clustering Plot Figure 5. Network visualization of reported sleep aspects. Nodes represent categories obtained from the content analysis of the questions: “What is for you a good night of sleep?” and “What is for you a good sleep quality?” Network Plot is the result of the application of the estimator EBICglasso to the data, which cluster the variables according to the appropriate test. The Centrality and Clustering measures are also important. Regarding the centrality measures there is: betweness, which provides information related to the nodes that have the highest number of shortest paths; closeness corresponds to the inverse of the sum of all shortest paths from the node of interest to all other nodes; and the degree is the sum of the absolute input weights of that node. From the network plot, it is possible to observe that the sleep quality (GSQ) parameters “characteristics”, “continuity”, “duration” and “final wake” are clustered together with a strong association between them. A cluster between sleep quality and sleep night (GSN) parameters is also observed: “GSQ_onset”-“GSN_room”-“GSQ_room”-“GSN_continuity”. In terms of centrality measures, the betweeness of “GSQ_characteristics” and “GSN_onset” is relatively high when 119 compared to other nodes, which means that there are more shortest paths passing through these two nodes, than to any other node (i.e. it is easier to traverse from other nodes to “GSQ_characteristics” and “GSN_onset”). In general, a higher centrality measure indicates that the node is more central to the network. When exploring the network analysis only for the question regarding the meaning of good sleep quality, results indicate that the parameters “final wake” and “sleep continuity” seem to be the most central for the obtained network. In fact, in the formed cluster (see network plot) the parameters sleep “continuity” and “final wake” are the ones with the stronger association. The parameter sleep “characteristics” appeared also as important being clustered to “final wake” and sleep “duration” and, in a weaker association, to sleep “continuity” (Figure 6). Network Plot Centrality Plot Clustering Plot Figure 6. Network visualization of reported sleep aspects. Nodes represent categories obtained from the content analysis of the question: “What is for you a good sleep quality?” Network Plot is the result of the application of the estimator EBICglasso to the data, which cluster the variables according to the appropriate test. The Centrality and Clustering measures are also important. Centrality measures: betweness, which provides information related to the nodes that have the highest number of shortest paths; closeness corresponds to the inverse of the sum of all shortest paths from the node of interest to all other nodes; and the degree is the sum of the absolute input weights of that node. 120 Stability of the definitions provided by the participants To determine the stability of the “good sleep quality” concept, n=53 participants were re-evaluated one year later. Of these, 52% used the same parameters to describe a good sleep quality compared to the year before (Figure 7). Table 3 shows the changes that occurred from moment 1 to moment 2 in the parameters used to describe a good sleep quality. Figure 7. Graphic representation of the results from the longitudinal qualitative analysis to the meaning of “good sleep quality” and the frequency of individuals that use the same or different parameters in its description. The figure on top represent the results from the qualitative analysis, and the bottom plot the frequency in the response. The light yellow color concerns the first moment of evaluation and the dark yellow color represents the second moment of assessment, one year later. Numbers from 1 to 7 concern the units of analysis that resulted from the qualitative analysis: 1-Sleep duration; 2-Sleep continuity; 3-Sleep onset; 4-Sleep characteristics; 5-Room environment; 6-Final wake up and rise; 7-Next day performance. Each column represents one study participant. Table 3. Observed changes in the parameters reported to define a good sleep quality from moment 1 to moment 2. Moment 1 Moment 2 Sleep characteristics Final Wake and Rise Sleep onset; Sleep characteristics Final wake and rise Sleep characteristics; Final wake and rise Final wake and rise Sleep continuity; Sleep characteristics; Final wake and rise Sleep continuity; Final wake and rise Sleep continuity Final wake and rise Sleep continuity; Sleep onset; Sleep characteristics Sleep characteristics Sleep characteristics Sleep onset Sleep characteristics Sleep duration; Sleep characteristics Sleep duration; Sleep continuity Sleep duration Sleep characteristics Sleep continuity; Sleep characteristics Sleep duration; Sleep continuity Sleep characteristics Sleep continuity; Sleep characteristics Sleep continuity; Room environment Sleep continuity Sleep characteristics Sleep continuity; Room environment Sleep continuity Sleep continuity Sleep duration Sleep duration; Sleep onset Sleep duration Sleep characteristics Sleep duration 1 2 3 4 5 6 7 127 when assessed one year later, changed the parameters used to describe each of the questions, possibly suggesting that despite discriminating between questions, the distribution of the parameters might be flexible. Interestingly, considering the meaning of a good sleep quality, the most reported parameters in our study (“sleep continuity”, “sleep characteristics” and “final wake”) are congruent to previous studies’ results, even with those that used an inter-individual approach (Akerstedt et al., 1994; Domino, Blair, & Bridges, 1984; Webb, Bonnet, & Blume, 1976). The use of network analysis to estimate the relation between all variables directly, instead of trying to reduce the structure of the variables to their shared information (latent variable modeling) is a relatively new and promising method for modeling interactions between large numbers of variables (Epskamp, Borsboom, & Fried, 2017). Interestingly, while in the first model all parameters were considered together (from the meaning of a good sleep quality but also from the meaning of a good night sleep), “sleep onset” was a central parameter of the model (also observed in Goelema and colleagues study (2018)). On the other hand, when considering only the parameters obtained in the sleep quality analysis, results showed not only that “sleep continuity” and “final wake” were central nodes to the obtained network, but that there was also a weak association between “final wake” and “next day performance”. This suggests that it may be important a more broad approach to the sleep quality meaning in order to properly cover all the important aspects of it. In a recent study from Goelema and colleagues, their results indicated “daytime functioning”, “interruptions during the night” and “before bed state” as major aspects of sleep experience for lay people (Goelema et al., 2018). To a certain degree, our results are similar. Nonetheless, a direct comparison is not possible since they were less restrictive in what they consider for each category. For example, in the category “sleep characteristics”, “other sleep parameters such as sleep onset latency, deep sleep and sleep duration” were considered; while, here we consider all the subjective aspects used to describe sleep (e.g. good sleep, restful sleep or components like with or without dreams or/and nightmares). Overall, it appears that individuals judge their sleep quality retrospectively considering both their memory of nighttime sleep and their experience during the day (Ramlee et al., 2018; Ramlee, et al., 2017), specifically their feelings upon waking. Furthermore, when conceiving sleep quality judgement as a decision making progress, daytime functioning was more important for people to judge their sleep experience than the actual night of sleep itself (Ramlee et al., 2018; Ramlee et 128 al., 2017). These -feelings upon waking and the evaluation of their mood and daytime performanceresonate with previous work that suggest a significant role of daytime impairments in the genesis of insomnia complaints (Ramlee et al., 2017). Considering that “sleep continuity” was the most mentioned category by our participants, here the progress of the night is of focus, which might also help explain feelings upon the final wake and its importance for our population. In fact, it was quite interesting to observe that often participants referred dreaming as indicator of poor sleep quality and that this trend has also been observed in other studies (Akerstedt et al., 1994; Goodenough, 1978). Those studies showed that the report of having no or few dreams was associated to fewer awakenings and, consequently, to the perception of dreaming as a proxy to sleep interruption and, consequently, as a signal of poor sleep. Furthermore, in the study from Goelema and colleagues (2008), participants indicated that their state of mind before bedtime was an import factor for the sleep experience, which has also been suggested by other studies (Åkerstedt et al., 2012; Eliasson et al., 2010). Interestingly, they pointed the link between categories, as stress before going to bed, which can promote longer sleep onset latency, more awakenings during the night and a shorter sleep duration (Akerstedt, 2006; Sadeh et al., 2004). Consequently, the individual may feel tired during the day. With the present work, we extend on Goelema and colleagues (2018) and further explore differences in sleep quality conceptualization by determining the association between different measures of sleep quality and not only on PSQI subdomains, but also by exploring the determinants of sleep quality. Specifically, considering the association with PSQI subdomains, results showed that the categorization does not match with the questions of the PSQI. For example, PSQI questions relate more to the extremes of daytime functioning (e.g. ‘During the past month, how often have you had trouble staying awake while driving, eating meals, or engaging in social activity?’), which does not always apply. For instance, while it is possible that people experience some degree of dysfunction during the day because of tiredness or not feeling refreshed, it is also possible that they do not experience problems with staying awake (Goelema et al., 2018). This point highlights the importance of a more comprehensive measure of sleep quality that can translate a continuous between health and disease. Following in the same line are the results concerning the association between self-reported measures of sleep quality (qualitative and quantitative) and PSQI (the standard measure for subjective sleep quality). A moderate association was observed, which made it clear for the need of a more comprehensive self-reported measure, because even though all 129 these measures assess the same construct, they are likely reflecting different aspects of sleep quality. Overall, it is clear that sleep quality has to be defined in a comprehensive manner. With this work it became clear that sleep quality is more than just sleep quantity, timing, simple stage structure, occurrence of pathologic events (Krystal & Edinger, 2008) or even one’s satisfaction with his/her sleep (Ohayon et al., 2017). The need of a tool that can translate not only a little of each parameter used in the description of sleep quality, but that can also consider some of the known determinants of sleep is of great interest. Furthermore, with this work we also showed that the meaning of a good sleep quality does not differ with age. Nonetheless, further research is needed not only comparing different cultural groups with different sleep patterns and contexts, but exploring factors such as circadian rhythm, day-to-day variability, use of sleep medication, social conventions (e.g.,weekday/weekend distinction) or weather (e.g., availability of sunshine). The understanding of the complex dynamic of perception and self-reported measures across the lifespan will cast some light in how we can obtain more reliable information in the clinic that can result in more efficient and adequate therapies. It was beyond the work scope to develop an index based in this qualitative definition of sleep quality. However, it should certainly be the next step and an approach similar to Akersted and colleagues (1994) can be followed. After their qualitative definition they asked individuals to rate each of the parameters and their overall sleep quality. Their results showed that sleep was rated "very good" if it contained 4% or less of waking, whereas "very poor" sleep corresponded to 24% or more of waking; it seemed that four minutes constituted a "very easy" sleep onset, in contrast, with "very difficult" that corresponded to a sleep latency of 100 min or more. Furthermore, "calmness of sleep" appeared maximal at 0.1 awakenings per hour or less and minimal at 0.5 awakenings per hour or more (Akerstedt et al., 1994) and they suggested that their results could be used as guidelines for qualitative interpretation of quantitative (but subjective) sleep parameters. Some methodological aspects should be addressed. Given the heterogeneity in studies that focus on the subjective components of sleep quality, for the present study we aimed to focus on this subjective dimension without explore objective components of sleep quality. To our knowledge, it is the first time that the same study determines the subjective meaning of sleep quality using not only a qualitative approach but also a network analysis of the data, exploring the association between different self-reported measures as well as the determinants of sleep quality considering the different measures. The choice to freely let the participants answer what 130 they consider to be the meaning of a good sleep quality was supported by the large number of participants, which confered some advantage to the approach given that having a large sample of individuals operationalizing the concept of a good sleep quality, provide less bias to the subjective interpretation of the phenomenon. Nonetheless, in future studies, it can be of interest to apply an automated analysis of free speech (Bedi et al., 2015) or even machine learning techniques. With the present work we overcame some limitations of previous studies, namely the reduced number of participants and the lack of information about the stability of the definition of good sleep quality. These results can guide future research to provide suitable psychometric measures for normal sleepers, as well as the design of sleep data visualization applications in the context of health self-monitoring. Conclusion The present works contributes to the literature by providing information in a still highly debatable question: what is considered a “good sleep quality” for the general population. We were able to provide qualitative information that corroborates previous results with smaller sample sizes, but we also add the possibility of studying quantitatively some of the aspects that remained open from previous studies, namely what happens to the meaning of a good sleep quality with age and what are the weighs of mood and personality as determinants of the answer provided but also the selfrating of sleep quality. 131 Conflict of interest statement The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Contributions LA, NS and NCS conceived the study. LA recruited the participants and collected the data with the help of Alexandra Sousa (a medical student). LA analyzed the data and wrote the first draft of the manuscript. All authors contributed to the final manuscript. Acknowledgments This work was developed under the scope of the projects NORTE-01-0145-FEDER-000013, supported by the Northern Portugal Regional Operational Programme (NORTE 2020), under the Portugal 2020 Partnership Agreement, through the European Regional Development Fund (FEDER). It has also been funded by FEDER funds, through the Competitiveness Factors Operational Programme (COMPETE), and by National funds, through the Foundation for Science and Technology (FCT), under the scope of the project POCI-01-0145-FEDER-007038. LA was supported by a FCT PhD scholarship with the SFRH/BD/101398/2014 and NCS was supported by a post-doc fellowship under the projet “CODIGOMAIS: Creación de un Ecosistema Transfronterizo de Innovación en Salud” (Interreg Espana-Portugal, Programa de Cooperación Interreg V A Espana-Portugal (POCTEP 2012-2020); Project reference: 0227). The authors would like to acknowledge Alexandra Sousa’s contribution in participants’ recruitment and data collection, and Teresa Costa Castanho for the collaboration in the final step of the content analysis. 132 References Akerstedt, T., Hume, K., Minors, D., & Waterhouse, J. (1994). The subjective meaning of good sleep, an intraindividual approach using the Karolinska Sleep Diary. Perceptual and Motor Skills , 79 (1 Pt 1), 287–296. https://doi.org/10.2466/pms.1994.79.1.287 Åkerstedt, T., Orsini, N., Petersen, H., Axelsson, J., Lekander, M., & Kecklund, G. (2012). Predicting sleep quality from stress and prior sleep--a study of day-to-day covariation across six weeks. Sleep Medicine , 13 (6), 674–679. https://doi.org/10.1016/j.sleep.2011.12.013 Akerstedt, Torbjörn. (2006). Psychosocial stress and impaired sleep. Scandinavian Journal of Work, Environment & Health , 32 (6), 493–501. Bedi, G., Carrillo, F., Cecchi, G. A., Slezak, D. F., Sigman, M., Mota, N. B., … Corcoran, C. M. (2015). Automated analysis of free speech predicts psychosis onset in high-risk youths. Npj Schizophrenia , 1 , 15030. https://doi.org/10.1038/npjschz.2015.30 Buysse, D. J., Reynolds, C. F., Monk, T. H., Berman, S. R., & Kupfer, D. J. (1989). The Pittsburgh sleep quality index: A new instrument for psychiatric practice and research. Psychiatry Research , 28 (2), 193–213. https://doi.org/10.1016/0165-1781(89)90047-4 Del Rio João, K. A., Becker, N. B., de Neves Jesus, S., & Isabel Santos Martins, R. (2017). Validation of the Portuguese version of the Pittsburgh Sleep Quality Index (PSQI-PT). Psychiatry Research , 247 , 225–229. https://doi.org/10.1016/j.psychres.2016.11.042 Domino, G., Blair, G., & Bridges, A. (1984). Subjective Assessment of Sleep by Sleep Questionnaire. Perceptual and Motor Skills , 59 (1), 163–170. https://doi.org/10.2466/pms.1984.59.1.163 Eliasson, A. H., Kashani, M., Mayhew, M., Ude, A., Hoffman, J., & Vernalis, M. (2010). Reducing Perceived Stress Improves Sleep Quality: A Longitudinal Outcomes Study. CHEST , 138 (4), 913A. https://doi.org/10.1378/chest.10417 Erlingsson, C., & Brysiewicz, P. (2017). A hands-on guide to doing content analysis. African Journal of Emergency Medicine , 7 (3), 93–99. https://doi.org/10.1016/j.afjem.2017.08.001 Goelema, M. S., de Bruijn, R., Overeem, S., Møst, E., Haakma, R., & Markopoulos, P. (2018). Conceptions of sleep experience: a layman perspective. BMC Research Notes , 11 . https://doi.org/10.1186/s13104-018-3584-2 133 Harvey, A. G., Stinson, K., Whitaker, K. L., Moskovitz, D., & Virk, H. (2008). The Subjective Meaning of Sleep Quality: A Comparison of Individuals with and without Insomnia. Sleep , 31 (3), 383–393. Johns, M. W. (1991). A new method for measuring daytime sleepiness: the Epworth sleepiness scale. Sleep , 14 (6), 540–545. Krystal, A. D., & Edinger, J. D. (2008). Measuring sleep quality. Sleep Medicine , 9 , S10–S17. https://doi.org/10.1016/S1389-9457(08)70011-X Lovibond, S. H., & Lovibond, P. F. (1996). Manual for the depression anxiety stress scales . Obtido de https://trove.nla.gov.au/version/38629420 Mander, B. A., Winer, J. R., & Walker, M. P. (2017). Sleep and Human Aging. Neuron , 94 (1), 19– 36. https://doi.org/10.1016/j.neuron.2017.02.004 Moul, D. E., Nofzinger, E. A., Pilkonis, P. A., Houck, P. R., Miewald, J. M., & Buysse, D. J. (2002). Symptom reports in severe chronic insomnia. Sleep , 25 (5), 553–563. Ohayon, M., Wickwire, E. M., Hirshkowitz, M., Albert, S. M., Avidan, A., Daly, F. J., … Vitiello, M. V. (2017). National Sleep Foundation’s sleep quality recommendations: first report. Sleep Health: Journal of the National Sleep Foundation , 3 (1), 6–19. https://doi.org/10.1016/j.sleh.2016.11.006 Pais-Ribeiro, J. L., Honrado, A., & Leal, I. (2004). CONTRIBUIÇÃO PARA O ESTUDO DA ADAPTAÇÃO PORTUGUESA DAS ESCALAS DE ANSIEDADE, DEPRESSÃO E STRESS (EADS) DE 21 ITENS DE LOVIBOND E LOVIBOND . 11. Ramlee, F., Afolalu, E. F., & Tang, N. K. Y. (2018). Do People With Chronic Pain Judge Their Sleep Differently? A Qualitative Study. Behavioral Sleep Medicine , 16 (3), 259–271. https://doi.org/10.1080/15402002.2016.1188393 Ramlee, F., Sanborn, A. N., & Tang, N. K. Y. (2017). What Sways People’s Judgment of Sleep Quality? A Quantitative Choice-Making Study With Good and Poor Sleepers. Sleep , 40 (7). https://doi.org/10.1093/sleep/zsx091 Rosa, R. R., & Bonnet, M. H. (2000). Reported Chronic Insomnia Is Independent of Poor Sleep as Measured by Electroencephalography. Psychosomatic Medicine , 62 (4), 474. Rosipal, R., Lewandowski, A., & Dorffner, G. (2013). In search of objective components for sleep quality indexing in normal sleep. Biological Psychology , 94 (1), 210–220. https://doi.org/10.1016/j.biopsycho.2013.05.014 134 Sadeh, A., Keinan, G., & Daon, K. (2004). Effects of stress on sleep: the moderating role of coping style. Health Psychology: Official Journal of the Division of Health Psychology, American Psychological Association , 23 (5), 542–545. https://doi.org/10.1037/02786133.23.5.542 Unruh, M. L., Redline, S., An, M.-W., Buysse, D. J., Nieto, F. J., Yeh, J.-L., & Newman, A. B. (2008). Subjective and Objective Sleep Quality and Aging in the Sleep Heart Health Study. Journal of the American Geriatrics Society , 56 (7), 1218–1227. https://doi.org/10.1111/j.15325415.2008.01755.x Webb, W. B., Bonnet, M., & Blume, G. (1976). A Post-Sleep Inventory. Perceptual and Motor Skills , 43 (3), 987–993. https://doi.org/10.2466/pms.1976.43.3.987 135 Appendix 1 Quantitative self-report scale 0 to 100 threshold determination for a poor sleep quality Methods The first step was to invert (direct inversion) the scale in order to obtain the same direction as in PSQI – higher scores indicating poorer sleep quality. This step is crucial given that the standard measure considered for the receiver operating curve (ROC) method was PSQI, a standard measure of subjective sleep quality. Results The results from ROC are presented in Figure S1. Overall, the analysis demonstrated satisfactory discriminant validity with the area under the curve pointing for a threshold score of equal or below 27.5 for good sleep quality [75.8% (95% CI: 69–82.5%)]. Figure S1. ROC curve of the quantitative (0-100) self-report scale. The area under the curve was 75.8% (95% CI: 69– 82.5%). 136 CHAPTER IV A short report on week-weekend variability of self-reported sleep patterns, routines and quality: preliminary results Liliana Amorim, Nuno Sousa, Nadine Correia Santos (In preparation) 143 is observed between sex, age and sleep latency (F(2, 290)Week=7.90, p<0.001, R2 adj=0.045; F(2, 290)weekend=8.52, p<0.001; R2 adj=0.049), both on week and weekend days. Figure 2. Characterization of the Rest Patterns of a sample of community-dwellers from 18 to 86 yrs on week and weekend days, by age and sex (n=307). Sleep routines at bedtime where also explored. On this, results show that during weekdays, 52.6% of the participants reported some type of activity when going to bed. This percentage slightly increased during weekend days (54.67%) with individuals spending also a little more time in those (Figure 3). In fact, considering only the participants that reported spending time in, at least, one of 144 these routines, results indicate that the total amount of time ranged from 5 to 210 minutes. Participants were engaged for more time in bedtime activities during weekends (Meanweek= 46.68 minutes vs Meanweekend= 54.38 minutes; t(147)= 3.106, p=0.002, R2=0.062). Figure 3. Characterization of the Rest Patterns of a sample of community-dwellers from 18 to 86 years on weekdays and weekend, by age (n=307). When testing the influence of age and sex, results show that with age the time spent in bedtime activities increase both in week and weekend days (Fweek= 17.56, p<0.001; Fweekend= 9.44, p=0.0025), and differences regarding men and women were only observed for weekdays, with men reporting more time in bedtime activities when at younger age, and women at higher ages. Exploratory longitudinal study and determination of week-weekend subjective sleep quality To determine how sleep patterns, routines and quality evolve throughout one year, an exploratory longitudinal study was developed. Results show only a statistically significant difference for bedtime between Moment 0 and Moment 1 (Z=199.5, p=0.003). Furthermore, the differences between week and weekend days were maintained in Moment 1, similarly to Moment 0 (Table 3). 145 Table 3. Longitudinal characterization of sleep patterns characterization of the subsampled cohort (n=53). Moment 0 Moment 1* Week (median, IQR) Weekend (median, IQR) Week (median, IQR) Weekend (median, IQR) Bedtime 23.0 (1.20) 23.5 (1.60) 23.0 (1.60) 23.3 (1.6) Light off 23.5 (1.20) 23.8 (1.50) 23.0 (1.50) 24.0 (2.0) Sleep latency 10 (10) 10 (12.5) 15min (12.5) 10min (10.0) Wake up time 7.0 (0.67) 8.5 (0.75) 07:00 (0.75) 8:00 (1.63) Time to rise after waking up 8.75 (10.0) 15 (7.5) 10min (7.5) 15min (15.0) Sleep duration 7 (1.50) 8 (2.0) 7h (2.0) 8h (2.0) Alarm clock use# 35 (71.4%) 11 (22%) 31 (69.8%) 11 (23.4%) * Moment 0 – occurred 12 months after Moment 0. # n (%) Our participants were also asked about the variation of their sleep quality in week and weekend days. A statistically significant difference was found, with reported sleep quality being better in weekends (Figure 4). Figure 4. Subjective sleep quality on week and weekend days. Interestingly, Saturday was the day of the week most reported has the better night sleep and Sunday the worse night in terms of sleep quality. Factors provided by participants as related to this was worrying about work for a worse night on Sunday as well as becoming alone (children return to rent home for university classes during the week) and exercise or off day as reason for the better sleep quality night (Figure 4). Week Wekends 0 20 40 60 80 100 Score Sleep Quality Self-Rating 146 Discussion In our 24-h society, the demands of the different roles that each individual performs can promote unhealthy lifestyles, particularly in what concerns to sleep habits and routines, which can ultimately result in sleep disruption and deprivation. While sleep disturbances have been shown to affect a wide variety of biological systems and processes (Medic et al., 2017; Potter et al., 2016; Wickens et al., 2015; Finan et al., 2015), irregularities in sleep schedule have been mostly associated to poor subjective sleep quality (Monk et al., 2003). Interestingly, these irregularities in sleep scheduling can often be part of a coping strategy to deal with sleep loss. For instance, it is not uncommon for the the individual to engage in daytime napping or to extend nighttime sleep during weekends or other periods free from social or work obligation, with the purpose of obtaining a few extra hours of sleep. Thus, studying week and weekend variation of sleep can provide valuable information to develop new strategies and models to overcome some of the challenges that are still associated to sleep complaints and disturbances. Differences between week and weekend days have been mostly addressed in adolescents and young adults (Hasler et al., 2012; Gradisar et al., 2008). However, as we can see with the present results, this variation also occurs across the adult lifespan, although less proeminently at older ages. In Portugal, very few epidemiological studies have focused on the sleep topic, and the last one dates from more than a decade ago. Thus, this work enabled a first characterization of sleep patterns and quality in a sample of adults within its normative ageing process. Overall, differences were found between sexes and across ages in what concerns sleep patterns and routines. Furthermore, the mean difference between weekend and weekdays was of 1 hour for bedtime, 5 minutes for sleep latency, 1,75 hours for waking time and 1,5 hours for the amount you sleep. In order to provide a better comprehension of how sleep habits and routines are in Portugal and what are the aspects that influence them, we should continuing exploring these data and how it evolves throughout time. Thus, it will be important to finish the follow-up assessments and to start building predictive models, but also, to expand the recruitment beyond the convinence sample, so that it will be possible to generalize the results to the adult Portuguese population. 147 Conflict of interest statement The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Acknowledgments The authors would like to acknowledge Alexandra Sousa for the help with the recruitment of the initial cohort of participants. Financial suport Financial support was provided by FEDER funds through the Operational Programme Competitiveness Factors - COMPETE and National Funds through FCT - Foundation for Science and Technology under the project POCI-01-0145-FEDER-007038, and by the project NORTE-010145-FEDER-000013, supported by Norte Portugal Regional Operational Programme (NORTE 2020), under the PORTUGAL 2020 Partnership Agreement, through the European Regional Development Fund (ERDF). The work was developed under the scope of the projects “SwitchBox” (European Commission, FP7; contract HEALTH-F2-2010-259772) and “TEMPO - Better mental health during ageing based on temporal prediction of individual brain ageing trajectories” (Fundação Calouste Gulbenkian; Contract grant number P-139977). LA was supported by the FCT PhD scholarship SFRH/BD/101398/2014. NCS was supported by a post-doc fellowship under the projet “CODIGOMAIS: Creación de un Ecosistema Transfronterizo de Innovación en Salud” (Interreg Espana-Portugal, Programa de Cooperación Interreg V A Espana-Portugal (POCTEP 20122020); Project reference: 0227). Contributions LA, NS and NCS designed the study. LA performed data collection and analysis. LA wrote the first draft. All authors wrote the final manuscript and gave input to the work. 148 References Asif, N., Iqbal, R., & Nazir, C. F. (2017). Human immune system during sleep. American Journal of Clinical and Experimental Immunology , 6 (6), 92–96. Brainard, J., Gobel, M., Scott, B., Koeppen, M., & Eckle, T. (2015). Health implications of disrupted circadian rhythms and the potential for daylight as therapy. Anesthesiology , 122 (5), 1170– 1175. https://doi.org/10.1097/ALN.0000000000000596 Dijk, D.-J., & Archer, S. N. (2010). PERIOD3, circadian phenotypes, and sleep homeostasis. Sleep Medicine Reviews , 14 (3), 151–160. https://doi.org/10.1016/j.smrv.2009.07.002 Gilman, N. V. (2018). Analysis for Science Librarians of the 2017 Nobel Prize in Physiology or Medicine: The Life and Work of Jeffrey C. Hall, Michael Rosbash, and Michael W. Young. Science & Technology Libraries . Obtido de https://www.tandfonline.com/doi/abs/10.1080/0194262X.2017.1412848 Mander, B. A., Winer, J. R., & Walker, M. P. (2017). Sleep and Human Aging. Neuron , 94 (1), 19– 36. https://doi.org/10.1016/j.neuron.2017.02.004 Ohayon, M. M. (2004). Interactions between sleep normative data and sociocultural characteristics in the elderly. Journal of Psychosomatic Research , 56 (5), 479–486. https://doi.org/10.1016/j.psychores.2004.04.365 Ohayon, M. M., & Paiva, T. (2005). Global sleep dissatisfaction for the assessment of insomnia severity in the general population of Portugal. Sleep Medicine , 6 (5), 435–441. https://doi.org/10.1016/j.sleep.2005.03.006 Simon, E. B., & Walker, M. P. (2018). Sleep loss causes social withdrawal and loneliness. Nature Communications , 9 (1), 3146. https://doi.org/10.1038/s41467-018-05377-0 Walker, M. P. (2008). Cognitive consequences of sleep and sleep loss. Sleep Medicine , 9 , S29– S34. https://doi.org/10.1016/S1389-9457(08)70014-5 Walker, M. P., & Stickgold, R. (2004). Sleep-dependent learning and memory consolidation. Neuron , 44 (1), 121–133. https://doi.org/10.1016/j.neuron.2004.08.031 Xie, L., Kang, H., Xu, Q., Chen, M. J., Liao, Y., Thiyagarajan, M., … Nedergaard, M. (2013). Sleep Drives Metabolite Clearance from the Adult Brain. Science (New York, N.Y.) , 342 (6156). https://doi.org/10.1126/science.1241224 149 CHAPTER V Poor sleep quality associates with decreased functional and structural brain connectivity in normative ageing: a MRI multimodal approach Liliana Amorim, Ricardo Magalhães, Ana Coelho, Pedro Silva Moreira, Carlos Portugal-Nunes, Teresa Costa Castanho, Paulo César Gonçalves Marques, Nuno Sousa, Nadine Correia Santos Published in Frontiers in Aging Neuroscience, DOI: https://doi.org/10.3389/fnagi.2018.00375 150 Poor sleep quality associates with decreased functional and structural brain connectivity in normative ageing: a MRI multimodal approach Authors: Liliana Amorim1,2,3, Ricardo Magalhães1,2,3, Ana Coelho1,2,3, Pedro Silva Moreira1,2,3, Carlos Portugal-Nunes1,2,3, Teresa Costa Castanho1,2,3, Paulo César Gonçalves Marques1,2,3, Nuno Sousa1,2,3, Nadine Correia Santos1,2,3 1Life and Health Sciences Research Institute (ICVS), School of Medicine, University of Minho, Campus Gualtar, 4710-057 Braga, Portugal. 2ICVS/3B’s - PT Government Associate Laboratory, Braga/Guimarães, Portugal. 3Clinical Academic Center – Braga, Braga, Portugal. *Corresponding author: Nadine Correia Santos, Life and Health Sciences Research Institute (ICVS), School of Medicine, University of Minho, Campus Gualtar, 4710-057 Braga, Portugal. Email: [email protected]. Phone: +351 253 604 806. 151 Abstract Sleep is a ubiquitous phenomenon, essential to the organism homeostasis. Notwithstanding, there has been an increasing concern with its disruption, not only within the context of pathological conditions, such as neurologic and psychiatric diseases, but also in health. In fact, sleep complaints are becoming particularly common, especially in middle-aged and older adults, which may suggest an underlying susceptibility to sleep quality loss and/or its consequences. Thus, a whole-brain modeling approach to study the shifts in the system can cast broader light on sleep quality mechanisms and its associated morbidities. Following this line, we sought to determine the association between the standard self-reported measure of sleep quality, the Pittsburgh Sleep Quality Index (PSQI) and brain correlates in a normative ageing cohort. To this purpose, 86 participants (age range 52 to 87 years) provided information regarding sociodemographic parameters, subjective sleep quality and associated psychological variables. A multimodal magnetic resonance imaging (MRI) approach was used, with whole-brain functional and structural connectomes being derived from resting-state functional connectivity (FC) and probabilistic white matter tractography (structural connectivity, SC). Brain regional volumes and white matter properties associations were also explored. Results show that poor sleep quality was associated with a decrease in FC and SC of distinct networks, overlapping in right superior temporal pole, left middle temporal and left inferior occipital regions. Age displayed important associations with volumetric changes in the cerebellum cortex and white matter, thalamus, hippocampus, right putamen, left supramarginal and left lingual regions. Overall, results suggest that not only the PSQI global score may act as a proxy of changes in FC/SC in middle-aged and older individuals, but also that the age-related regional volumetric changes may be associated to an adjustment of brain connectivity. These findings may also represent a step further in the comprehension of the role of sleep disturbance in disease, since the networks found share regions that have been shown to be affected in pathologies such as depression and Alzheimer’s disease. Keywords: Pittsburgh Sleep Quality Index; PSQI; MRI; Whole-brain modeling; Brain Connectivity; resting-state. 152 Introduction Sleep is a recurring and reversible neurobehavioral state that involves reduced responsiveness to external stimuli and is frequently accompanied by postural recumbence and behavioral quiescence (Carskadon and Dement, 2011). Propensity to sleep is determined by the interaction of the circadian rhythm (‘process C’), controlled by the suprachiasmatic nucleus, and a sleep homeostatic process (“process S”), that increasingly drives the need for sleep as a function of the time spent awake (Borbély, 1982; Borbély et al., 2016). For the individual to thrive and efficiently cope with the waking day demands, guidelines advise a human adult to sleep seven to nine hours every day (Hirshkowitz et al., 2015) and to have sleep continuity parameters (i.e. sleep latency, awakenings >5 minutes, wake after sleep onset and sleep efficiency) within a specific range (see Ohayon et al., 2017 for details). In fact, having a ‘good sleep’ ensures metabolic homeostasis, cerebral clearance, adequate immune function and overall good cognitive and mental status (Cirelli and Tononi, 2008; Freeman et al., 2017; Irwin, 2015; Shokri-Kojori et al., 2018; Stickgold, 2006; Tononi and Cirelli, 2006; Xie et al., 2013). However, factors such as inappropriate exposure to light or food, lifestyle schedules, work, or psychological morbidity can interfere with the appropriate timing and duration of the sleep/wake cycle, leading to wide-range adverse effects on health (Archer and Oster, 2015; Schwartz et al., 1999; Stickgold, 2006; Wulff et al., 2010). More so, age also emerges as a critical modifier of sleep-wake patterns, being responsible for a shorter overall sleep duration and increase in sleep fragmentation and fragility, mostly after the fifth decade of age (Mander et al., 2017). In addition, in an increasingly older population (United Nations, Department of Economic and Social Affairs, Population Division, 2013), sleep dissatisfaction is one of the most common complaints in primary care (Aikens and Rouse, 2005) contributing not only for a growth vulnerability to disease, but also for a considerable economic burden, consequence of the costs of sleep aids and work absenteeism (Hillman et al., 2006). In addition, the dyad sleep-depression is of relevant weight since not only sleep problems may underlie an increased risk for the middle-aged or older individuals to be depressed (Almeida and Pfaff, 2005), but are also a robust predictor of incident depression (Livingston et al., 1993; Mallon et al., 2000). In view of these associations and comorbidities, and because many neurological and psychiatric disorders share underlying brain network disturbances (Buckholtz and Meyer-Lindenberg, 2012; Deco and Kringelbach, 2014; Uhlhaas and Singer, 2012), it is vital to determine how one’s sleep complaints and perceptions affect neural circuitries and overall biological systems. On this, the