|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 SLEEP MISPERCEPTION AMONG COMMU NITY DWELLING OLDER ADULTS By DANIEL B. KAY A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2009
2 2009 Daniel B. Kay
3 To my family, especially my loving wi fe, thank you for your prayers and support
4 ACKNOWLEDGMENTS First, I want to acknowledge, Dr. Christina Mc Crae, for giving m e the opportunity to work and train under her mentorship. I would also lik e to thank my many mentors over the years that have helped me develop as a sleep theorist a nd researcher including Dr James Krueger for his inspirational ideas, Dr. Tadanobu Yasuda for his patience in teaching me the essence of sleep, Dr. Chris Davis for cutting me the ropes, and Dr. James Higley for his concern and confidence in me. I would like to acknowledge Dr. Meredeth Rowe for her involvement in compiling the database used in this study and for allowing me to explore my research with it. Special acknowledgement is also given to Joseph Dzierzwe ski for teaching me the statistical techniques used in this study. Special thanks is extended to Lisa and Marlin Kay and Janet and Rodney Hopkins for their good example, prayers, and love. It has been my pleasure to make you proud. Thanks is also extended to my wife, Janene Kay, for her patie nce, love, and support. Finally, for providing extra motivation, Leah, Corban, and Brice, I love you.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ............................................................................................................... 4LIST OF TABLES ...........................................................................................................................7LIST OF FIGURES .........................................................................................................................8ABSTRACT ...................................................................................................................... ...............9 CHAP TER 1 INTRODUCTION .................................................................................................................. 112 REVIEW OF THE LITERATURE ........................................................................................13Sleep Misperception in Older Adults ..................................................................................... 14Theoretical Perspectives of Sleep Misperception ................................................................... 15Sleep Misperception as Perceptu al Distortion or Deficit ................................................ 15Sleep Misperception as Exaggeration ............................................................................. 15Sleep Misperception as Psychopathology ....................................................................... 16Sleep Misperceptions as a Di stinct Sleep Disorder .........................................................18Sleep Misperception: The First and Ma intaining Symptom of Insomnia ....................... 19Sleep Misperception as a Graded Characteristic of Insomnia ......................................... 20Intraindividual Variabi lity Analysis (IIV) .............................................................................. 21Sleep Misperception in Context .............................................................................................. 24Mechanisms of Sleep Misperception ......................................................................................25Hyperarousal .................................................................................................................. .........25Cognitive Arousal ............................................................................................................ 26Physiological Arousal ......................................................................................................27CNS Hyperarousal ........................................................................................................... 27Sleep Misperception as Localized Sleep Deprivation ..................................................... 28Summary ....................................................................................................................... ..........323 STATEMENT OF THE PROBLEM ......................................................................................33Specific Aim 1 ........................................................................................................................33Hypothesis for Specific Aim 1 ........................................................................................ 33Specific Aim 2 ........................................................................................................................34Specific Aim 3 ........................................................................................................................35Specific Aim 4 ........................................................................................................................374 METHODS ....................................................................................................................... ......39Participants .................................................................................................................. ...........39Procedures .................................................................................................................... ...........40
6 Measures ...................................................................................................................... ...........41Objective Sleep Variables ...............................................................................................41Subjective Sleep Variables .............................................................................................. 42Sleep Misperception Variables (SOLsm and WASOsm) ................................................... 43Daytime Functioning Measure ........................................................................................ 43Beck Depression Inventory Second Edition (BDI-II) ...................................................43Demographics and Health Survey ...................................................................................43Data Analyses .........................................................................................................................44Specific Aim 1 .................................................................................................................44Specific Aim 2 .................................................................................................................45Specific Aim 3 .................................................................................................................45Specific Aim 4 .................................................................................................................485 RESULTS ....................................................................................................................... ........50Specific Aim 1 ........................................................................................................................50Specific Aim 2 ........................................................................................................................51Specific Aim 3 ........................................................................................................................52Multilevel Model for WASOsm ........................................................................................52Multilevel Model for SOLsm ............................................................................................53Specific Aim 4 ........................................................................................................................54Multicollinearity ............................................................................................................. .54Within-Person Multicollinearity ......................................................................................54Between-Person Multicollinearity ................................................................................... 54Total Sleep Time (TST o) ................................................................................................. 546 DISCUSSION .................................................................................................................... .....64Review of Findings .................................................................................................................64Aim 1 and 2 .....................................................................................................................64Aim 3 ...............................................................................................................................65Aim 4 ...............................................................................................................................66Summary of Results ................................................................................................................67Study Limitations ............................................................................................................. .......67A New Model of Sleep Misperception ................................................................................... 68Implications for Sleep Resear ch, Diagnoses, and Treatments ................................................ 71Treatment implications ...........................................................................................................73Future Directions ....................................................................................................................74APPENDIX A SLEEP DIARY .......................................................................................................................78B HEALTH SURVEY ...............................................................................................................79REFERENCE LIST .......................................................................................................................82BIOGRAPHICAL SKETCH .........................................................................................................95
7 LIST OF TABLES Table page 5-1Amount of withinand betw een-person variability ...........................................................565-2Amount of withinand between-person variability by complaint status. ..........................575-3Steps taken in building the WASOsm multilevel model ..................................................... 585-4The relationship between SOLsm and WASOsm .................................................................595-5Steps taken in building the WASOsm multilevel model ..................................................... 605-6The relationship between SOLsm and WASOsm .................................................................615-7Steps taken in building the TSTo multilevel model ........................................................... 625-8Sleep misperception variables predicting TSTo .................................................................63
8 LIST OF FIGURES Figure page 5-1 Relative amount of within-p erson variability co mpared to between-person variability after controlling for any linear or quadratic effects of tim e. .............................................. 51 5-2 Relative amount of within-p erson variability co mpared to between-person variability after controlling for any linear or quadratic effects of tim e.. ............................................. 52
9 Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science SLEEP MISPERCEPTION AMONG COMMUNITY DWELLING OLDER ADULTS By Daniel B. Kay May 2009 Chair: Christina S. McCrae Major: Psychology Late-life insomnia is a prevalent and seri ous health problem. Sleep misperception (SM), overestimating time spent awake while trying to sl eep, predicts insomnia onset and maintenance. Like insomnia, sleep misperception increases with age. However, resear ch on the longitudinal patterns and correlates of sleep misperception among older adults is needed. We posit that sleep misperception results from perceptual areas of the brain remaining awak e during global sleep, reflecting a form of localized sleep deprivation. Sleep misperception patterns may relate to a homeostatic sleep dysregulation, su ch as sleeping longer than usua l. In this study four specific aims that will be investigated in a sample of community dwelling older adults. Aim 1 will determine the amount of w ithinto between-person va riability in SM that occurs at sleep onset latency (SOLsm) and that occurs during wake time after sleep onset (WASOsm). Aim 2 will determine the amount of withinto between-perso n variability in SOLsm and WASOsm by sleep complaint status. Aim 3 will be to investigate the relationship between night-to-night fluctuations in SOLsm and WASOsm. Finally, aim 4 will determine the intraindividual relationship between SOLsm and WASOsm to objective TST (TSTo). A sample of 103 commun ity dwelling older adults (Mage = 72.81, SD = 7.12) wore an Actiwatch-L (24hrs/day/2weeks) and concurrently completed sleep diaries. Daily values for actigra phically-measured sleep onset latency (SOL) and
10 wake time after sleep onset (WASO) were subtract ed from respective diary reports to calculate daily SM for SOL and WASO. Intraindividual vari ability analyses (IIV) and multilevel modeling (MLM) revealed-1) within-p ersons, nights that WASOsm was greater than usual, SOLsm was less than usual, = -0.1, t (87.59) = -2.01, p < .05. Interestingly, between-p ersons, those with greater SOLsm had greater WASOsm, = 0.21, t(98.09) = 3.98, p < .01; 3) within-persons. There was a nonsignificant trend using SOLsm to predict WASOsm. ( = -0.10, t(90.90) = -1.80, p = .08); 4) nights with increased SOLsm were related to longer TSTo, = 0.32, t(67.74) = 5.02, p < .01 while nights with increased WASOsm were related to shorter TSTo, = -0.15, t(49.77) = -2.47, p < .05. Between-persons, only SOLsm was related to longer TSTo. Sleep misperception is ubiquitous but highly variable among older adults. I ndividuals with higher average SOLsm had higher average WASOsm; however, days individuals had higher WASOsm predicted lower SOLsm. A homeostatic sleep rebound model of sleep misperception may explain the finding that increased TSTo occurred on days that SOLsm was greater than usual. Unders tanding the variable, but predictive patterns of SM among older adults may lead new preventative and therapeutic insomnia interventions.
11 CHAPTER 1 INTRODUCTION Various constructs and research variables of sleep m isperception have been developed. Usually, sleep misperception is conceptualized as an individuals tendenc y to report objectively defined sleep as wakefulness (Adam, Tomeny, & Oswald) and to over-estimating the time spent falling asleep and underestimating total sleep time Sleep misperception is common to insomnia and most studies on sleep misperception are based on a single night observation or an average of several nights of sleep in this patient population. These studies have yielded valuable information about the relationship between sleep mispercep tion and insomnia; however, methodologies used in these studies are commonly based on scientifically untested assumptions. Herein, these assumptions are reviewed. This literature review also covers the scie ntific evidence supporting various perspectives on sleep misperception. Ultimately, this study seeks to expand a specific theory that argues that sleep misperception is a transitional state of consciousness on a continuum between wake and sleep that represents a symptom of a clinically significant sleep problem. Four secondary analyses were conducte d on data collected from a sample of older adults with the aim of better understand sleep mi sperception. The first two analyses challenge the common assumptions that sleep misperception is a c onsistent pattern unique to individuals with sleep complaint. Scientifical ly, this study seeks to provide a clearer picture of sleep misperception with greater temporal resolution than can be obtained using traditional single night sleep recording and multiple night averaging tech niques alone. The third analysis explores the relationship between sleep misperceptions that occur during sleep onset latency or SOLsm, and sleep misperceptions that occur durin g wake time in the night or WASOsm. The final analysis explores the possibility that sleep misperception is related to a true sleep problem predicted by longer sleep time. Will this study will not be able to directly test the th eoretical involvement of
12 neurophysiology to sleep mispercep tion, neurophysiologic data fr om previous studies were paramount in driving this investigati on and the interpretation of this study.
13 CHAPTER 2 REVIEW OF THE LITERATURE Broadly, sleep m isperception is an individual s tendency to perceive objectively defined sleep as wakefulness. Likewise, individuals th at misperceive sleep overestimate the time spent falling asleep, overestimate time spent awake duri ng the night, and/or underestimate total sleep time. Sleep misperception is determined by compari ng an individuals retros pective self-report of the sleep experience with an objective prospective measure (i.e., electroencephalograph or actigraphy) of the sleep state. While sleep misperception has been well studied, this phenomenon remains poorly understood. Research has shown significant relations hips between sleep misperception and mood, arousal, and general h ealth (Bonnet & Arand, 1997b; Carskadon et al., 1976; Chambers & Kim, 1993). In addition, sleep experts have long observed that sleep misperception is strongly associated with inso mnia (Frankel, Coursey, Buchbinder, & Snyder, 1976). Many posit sleep misperception acts as a primary pathway to insomnia onset and maintenance (Borkovec, 1982; Borkovec, La ne, & VanOot, 1981; Harvey, 2002a; Lundh & Broman, 2000; Perlis, Giles, Bootzin et al., 1997). To explain the role of sleep misperception in insomnia, cognitive theorists argue that sleep misperception leads to daytime protective behaviors that exacerbate arous al, thus leading to, and perpet uating, insomnia (Harvey, Tang, & Browning, 2005). The consequences of sleep misper ception in relationship to insomnia may be profound. Insomnia is among the most prevalent a nd costly health problems with social cost estimated between $92.5 and $107.5 billion each y ear (Stoller, 1994; Walsh & Engelhardt, 1999). In addition, insomnia is related to redu ced cognitive functioning, reduced performance, greater utilization of health services and more missed work days (Ohayon, Caulet, Priest, & Guilleminault, 1997). Moreover, the onset of inso mnia predisposes individuals to many other mental and physical health problems includi ng, depression, anxiety, pain and substance abuse
14 disorders (Ford & Kamerow, 1989). Considerin g sleep misperception may precipitate and maintain insomnia, research aimed at gaini ng knowledge about sleep misperception may be a fruitful avenue leading to new preventative and therapeutic interventions for insomnia. Sleep Misperception in Older Adults Adults over the age of 65, ofte n referred to as older adults, m ake up 13% of the total US population and have the highest ra tes of sleep disturbance of a ny age group. Older adults have more awakenings during the night, decreased sl eep efficiency, more variable nocturnal total sleep time, more alpha (light), less delta (deep ) sleep, more globala nd micro-arousal during sleep, and increased daytime napping (Nau, McCr ae, Cook, & Lichstein, 2005). Not surprisingly, older adults have higher rate s of insomnia complaints (K iley, 1999; Ohayon & Caulet, 1996) with greater associated severity and chronic ity (McCrae et al., 2003). Epidemiological studies estimate that the prevalence of insomnia is roughly 65% among older adults (Newman, Enright, Manolio, Haponik, & Wahl, 1997), thus twice as prevalent as occurs in younger populations (reviewed in Nau et al., 2005; Ohayon, 2002). In terestingly, with increased age subjective accounts become less reflective of objective measures of sleep. As is th e case with insomnia, sleep misperception increases with age. Additiona l research on the relationship between sleep misperception and the increase in sleep problems among older adults has been called for in the literature (Bonnet & Moore, 1982). Moreover, a better understanding of the relationship between objective and subjective measures in older ad ults is needed (Hoch et al., 1987). Sleep misperception research may yield valuable in formation beyond what can be gleaned from a direct relationship of either s ubjective or objective measures alone The current study seeks to fill this need by investigating the pa tterns and correlates of sleep misperception among older adults.
15 Theoretical Perspectives of Sleep Misperception Sleep Misperception as Perceptu a l Distortion or Deficit The phenomenon of sleep misperception re mains poorly understood (Tang & Harvey, 2005), and a ubiquitously accepted definition or unifying theory of sleep misperception is lacking. Sleep misperception has been investig ated as 1. a tendency to exaggerate sleep difficulties, 2. a perceptual deficit or pathological distortion of reality related only to sleep, 3. an insomnia specific symptom, and 4. a categorical insomnia disorder subtype. Although these perspectives have never been c ohesively integrated, they may be compatible. Below, several methodological and theoretical appr oaches of sleep mi sperception research are considered. The goal of this review is to contextualize the opera tional definition of sleep misperception tested in this study. Sleep Misperception as Exaggeration People with insom nia are often stereot yped as having unbelievable or seemingly exaggerated claims about sleep le ngth or sleep difficulties. For ex ample, patients with insomnia often report that they only sleep a few hours each night while spending 8-11 hours trying to sleep in bed each night. Moreover, for an insomnia patie nt to claim less than 4 hours of sleep per night while electroencephalography (EEG) records normal sleep is not uncommon. Objective measures of sleep rarely support such claims and as a result, many physicians and sleep professionals have concluded th at sleep misperception represents the patients propensity to dramatize or exaggerate sleep difficulties (V anable, Aikens, Tadimeti, Caruana-Montaldo, & Mendelson, 2000). However, there is no empirical ev idence that sleep misp erceptions represents these patients tendency to lie or exaggerate. In fact, there is compelling evidence to the contrary. Perlis (2001) showed, for example, that the discrepancies between objective and subjective measures are not uniform across sleep paramete rs among insomnia patient. He argued that if
16 sleep misperception were a negative bias following poor sleep the patient would complain about SOL, WASO, total sleep time (TST), sleep quality, sleep efficiency, etc. However, this is not what was found. In fact, patients with sleep mi sperception typically misp erceived only one or two aspects of their sleep. Sleep Misperception as Psychopathology Rather than viewing the tendency to ove r-es timate SOL and WASO as a form of mis reporting, most sleep experts consider this tendency a type of mis perception. Sleep misperception is most commonly described as a per ceptual deficit/distortion, such as an inability to estimate time or lack of ability to perceive the passing of time as others do. Though tempting to believe, sleep misperception as a global defi cit in time perception ha s no scientific support (Rioux, Tremblay, & Bastien, 2006). Moreover, t hough ubiquitously maintained (e.g., Borkovec, 1982; Borkovec, Grayson, O'Brien, & Weerts, 1979; Perlis, Giles, Mendelson, Bootzin, & Wyatt, 1997; Perlis, Merica, Smith, & Giles, 20 01; Tang & Harvey, 2004; Vanable et al., 2000), there is little evidence th at sleep misperception represents a per ceptual distortion of time specific to the sleep experience. One study attempting to identify such a deficit was unable to induce a state specific distortion of th e passing of time in insomnia patients (Tang & Harvey, 2005). Regrettably, this null finding has not dampened th e fervor that such a perceptual deficit exists. As is indicated by its moniker, sleep mi sperception is regarded as a psychological distortion of reality, that is, that individuals who experience sleep misperception perceive the same EEG defined state differently than those who do not misperceive. An early study on sleep misperception purported to have su pported this claim showing that the same EEG defined global state of sleep was perceived to be lighter by sleep complainers as compared to controls (Mendelson, Garnett, Gillin, & Weingartne r, 1984). Unfortunately, given that global psychophysiological measures cannot tell the difference between joy and anger, it seems odd that
17 scientists should claim that the mis is associated with the patient perception and not in the researchers measure. Indeed, the American Sleep Disorders Association (now called the American Academy of Sleep Medicine) pointed out that standard polysomnographic (PSG) measures used in research and clinical assessment are inadequate at determining altered physiological states in insomnia (Reite, Buysse, Reynolds, & Mendelson, 1995). It is presumptive, therefore, for scientists to go as far as to say that peopl e who experience sleep misperception suffer from difficulty in rec ognizing their own state of consciousness (Mendelson, James, Garnett, Sack, & Rosentha l, 1986, p. 268). A recent term that has emerged in the sleep literature that highlights this overconf idence in our assessment tools is sleep quality misperception(e.g., Neu, Mairesse, Hoffmann, Dris Lambrecht, et al., 2007). This term suggests that there are some i ndividuals who are incapable of determining how refreshing their sleep is and that a PSG machine knows better than the patient. Researchers ha ve gone as far as to teach individuals with sleep misperception how to perceive their sleep experience (Downey & Bonnet, 1992). Although this technique may inform patients on how the clinician desires him or her to report the sleep experience, the sleep expe rience of these patients may not have actually changed. There are at least four major reasons why sleep misperception should not be considered a psychopathological distortion of reality. First, sleep misperception occurs spontaneously (Sewitch, 1984) in 40-60 % of h ealthy sleepers (reviewed in Atta rian, 2007). On this point, the claim has been made that the literature show s that good sleepers tend to correctly estimate or slightly underestimate SOL (Smith & Trinder, 2 000). However, this is a qualitative assessment and is not overly descriptive. At best, the liter ature suggests that sleep misperception is more frequent and/or extreme among insomnia patients th an controls but not that it is absent from
18 controls (Perlis, Smith, Andrews, Orff, & Giles, 2001). Second, sleep misperception has been induced scientifically in norma l sleepers by increasing arousal through caffeine administration before sleep onset (Bonnet & Ar and, 1997b), an advanced phase shif t of the sleep schedule, and the manipulation of respiratory control to induce arousal (Smith & Trinder, 2000). Third, individuals with sleep misperce ption are not clinically psychopathological. As a group, they do not report elevated scores on the Minnesota Personality Inventory (MMPI-II) (Mendelson et al., 1986) nor do they have signifi cantly greater MMPI scores than normal sleepers (Bonnet & Arand, 1995). Forth, and most importantly, ther e is no contradictory evidence that sleep misperception reflects an accurate perception of an altered state and a true sleep disturbance commonly experienced by insomnia patients (Hauri & Olmstead, 1983). Sleep Misperceptions as a Distinct Sleep Disorder The International Classif ication of Sleep Diso rders-II (ICSD-II) includes a diagnosis of sleep state misperception as a distinct s ubgroup of insomnia. The ICSD-II outlines that individuals who meet crite ria for this sleep disorder (1) compla in of insomnia but lack significant EEG defined deficits that correspond to the comp laints and (2) do not have mental or physical ailment that might explain the misperception. Th e sub-categorization of sleep misperception is unfortunate, because most research on sleep misperception has focused on the diagnostic category rather than on the act ual phenomenon of sleep misper ception. This problem has been associated with sleep misperception almost from its discovery. Attempts to subcategorize individuals with sleep misperception within inso mnia have been made for decades and various diagnosis categories have been advocated, su ch as pseudoinsomnia, subjective insomnia, experimental insomnia, and sleep hypochrondriasi s. Pseudoinsomnia put forth by Borkovec et al. (1979) used sleep misperception as the sole feature of the diso rder. While subjective insomnia was based more on the assumption that everyone has the same sleep need and that if sleep length
19 was about average there is no real problem w ith the patients sleep. Whereas in 1988 it was argued by Trinder that, Although ove r-estimation of sleep disturba nce must occur in subjective insomnia, it does not appear to be uniquely asso ciated with the clinically identified condition. Therefore, its occurrence is irre levant to the identification, defin ition, or etiology of the disorder (Trinder, 1988, p. 91). The ICSD-II seems to have compromised between subjective insomnia and pseudoinsomnia. The presence of mispercepti on has become central to this diagnostic category in addition to the requirement that no global deficits be present (ICSD-II). By adding the construct of a mispercepti on to subjective insomnia, t hose advocating this diagnostic category have attempted to remove the complicati on of individual sleep n eed inherent with the former construct of subjective insomnia. However, the validity of sleep st ate misperception as a categorical sleep disorder has been questioned on theoretical and scie ntific ground (Dorsey & Bootzin, 1997; Edinger & Fins, 1995; P. J. Hauri & Wisbey, 1992; McCall & Edinger, 1992; Reynolds, Kupfer, Buysse, Coble, & Yeager, 199 1; Sugerman, Stern, & Walsh, 1985). Attempts to validate sleep misperception as a categorical subtype of insomnia are methodologically flawed (Salin-Pascual, Roehrs, Merlot ti, Zorick, & Roth, 1992) either limited to a single night of recording (e.g., Edinger et al ., 1996; Krystal, Edinger, W ohlgemuth, & Marsh, 2002; SalinPascual et al., 1992) or an average of several nights (e.g., Means, Edinger, Glenn, & Fins, 2003). The major problem with these approaches is di scussed in detail in a section below (see Intraindividual Variability Analysis (IIV)). Sleep Misperception: The First and Ma intaining Symptom of Inso mnia There is scientific evidence to support the idea that sleep mi sperception represents a mild or prodromic phase of insomnia that may later de velop into insomnia (Sal in-Pascual et al., 1992). Individuals with sleep mispercept ion lack objective indices of a sleep problem but have similar symptoms as insomnia patients, though less se vere (Bonnet & Arand, 1997b; Salin-Pascual et al.,
20 1992). The major problem with the idea that sl eep misperception is a prodromic phase of insomnia is that sleep misperception is so common among insomnia patients. How can sleep misperception be a prodromic phase of insomnia a nd a characteristic of even the most severe insomnia patients? One explanation is that indi viduals with sleep misperception may experience sleep onset similar to insomnia patients (H auri & Olmstead, 1983) but have a more isolated/localized sleep problem that predisposes them to a more severe insomnia. Indeed, sleep misperception seems to be on a continuum between normal sleep and chronic insomnia. Sleep misperception patients tend to having intermediate levels of vigilance (as measured by the multiple sleep latency test (MSLT)), short term memory problems, subjectively low vigor, and subjectively low estimates of sleep quality between normal and physiological insomnia patients (Bonnet & Arand, 1995). Sleep Misperception as a Graded Characteris tic of Insomnia There is substantial evidence that sleep mi sperception occurs in varying degrees of intensity and frequency in the vast majority of the general population but w ith increasing frequency or severity among insomnia patients (Edinger & Fins, 1995; Edinger et al., 2000; Fichten, Creti, Amsel, Bailes, & Libman, 2005; Krystal et al., 2002; Means et al., 2003; Schneider-Helmert & Kumar, 1995) Moreover, it has been argued that sleep misperception is a defining feature of insomnia regardless of degree (Bonnet & Arand, 1997a; P. Hauri & Olmstead, 1983; Reynolds et al., 1991; Salin-Pascual et al., 1992; Tang & Harvey, 2004; Trinder, 1988). Although overstretching the curr ent scientific evidence, many current sleep researchers and clinicians continue to uphold that sleep misperception oc curs constantly in a large portion of insomnia patients and is a valid characte rization of insomn ia (e.g., Edinger & Krystal, 2003).
21 Intraindividual Variability Analysis (IIV) The literature is conflicted as to whether sl eep m isperceptions follow a daily or occasional pattern and whether sleep mispercep tion is a defining feature of in somnia, a sub-classification of insomnia or distinct sleep disorder. Many of the discrepancies in findings and opinions related to sleep misperception are the consequence of inhe rent limitations to methodology of the vast majority of sleep misperception studies. Many stud ies are limited to a sing le night of laboratory PSG recording, and other studies obtain multiple nights of observations but then average over days and weeks. Both approaches are problem atic because, like objectiv e measures of sleep, sleep estimates are not stable over time (Edinger et al., 1997b; Edinger, Marsh, McCall, Erwin, & Lininger, 1991; Wohlgemuth, Edinger, Fi ns, & Sullivan, 1999); nor are objective and subjective measures systematically related over time (Edinger et al., 1991; Harvey, 2000; McCrae et al., 2005). Therefore, if sleep misp erception were to occur occasionally, single random night recordings are incapable of capturing sleep misperception in all insomnia participants, although all particip ants may experience sl eep misperception several times per week (McCall & Edinger, 1992). Researchers and clinicians attempting to glean information about slee p misperception from single night studied must use cau tion in interpreting the results. One larger clinic based study attempted to determine what percent of patient s have been diagnosed with sleep misperception (Coleman et al., 1982). It was determined that 9.2% of 1214 diagnosed insomnia patients met criteria for sleep state misperception. The diagnostic criterion for sleep state misperception is determined by a single night of PSG recording. T hus the number of indivi duals diagnosed with the sleep state misperception subtype of insomn ia may actually be the odds of someone with insomnia to experience sleep misperception during the night that the sl eep recording took place (i.e., 1:4). Other studies sugge sted that between 25% and 50% of insomnia cases can be
22 subsumed as sleep state misperception (D orsey & Bootzin, 1997; Mendelson, 1995c; SalinPascual et al., 1992; Sugerman et al., 1985). Agai n because these studies were based on a single night of sleep, these results could represent the od ds of insomnia patients to misperceive on any given night. Studies that aver age sleep perceptions over severa l days found that between and 34.4% and 38% of the sample could be subsumed as sleep state misperception (Hauri & Wisbey, 1992; Means et al., 2003). Because these studies av eraged over 3 and 6 days, respectively, they can be interpreted as the odds (i.e., 1:3) of so meone with insomnia experiencing misperception more often than not or at such an extreme degree on occasional nights to not be washed out by days sleep misperception did not occur. In sum, this disorder is based on clinical judgment, unsuccessful post-hoc experimental validation attemp ts, and a dearth of sc ientific evidence. The creation of another categorical disorder centered on sleep misperception effectively represents the construct of yet another invalid ICSD-II diag nostic category driven by clinical consensus rather than scientific evidence (Edinger et al ., 1996). For this reason, the current study will not limit its investigation of sleep misperception to individuals that meet criteria for the sleep disorder called sleep state misperception. Anothe r study separated individuals into objective and subjective insomnia patients base d on whether subjective complain ts were validated by a single night of PSG (Krystal et al., 2002). Only indi viduals who lacked objective measures of sleep problems experienced high beta ac tivity during sleep. This study may be interpreted as having shown that sleep misperception is less likely to occur on nights that objective sleep is poor or that measurement of sleep misperception on PSG measured poor sleep. Clinically, insomnia seems to be periodic, wi th several days of intense insomnia followed by nights of normal sleep (Drummond, Smith, Orff, Chengazi, & Perlis, 2004; Karacan, 1972). Scientifically, insomnia is not necessarily a constant sleep pattern from night-to-night. For this
23 reason, averaging several days to describe the sl eep of insomnia patients may be inappropriate. Likewise, it should not be assu med that averaging sleep misper ceptions across days is a valid indication of the patterns of i ndividuals sleep perceptions. For ex ample, combining overestimates of sleep difficulties with underestimates is bound to wash out occasional patterns of misperception. While averaging techniques undoubted ly have provided great er insight to sleep misperception, the more occasional or subtle patt erns of sleep misperception may be additionally important. Averaging techniques have been partic ularly useful in iden tifying individuals who consistently misperceive, that is, those i ndividuals who fall on the extremes of sleep misperception behavior. In fact, th is is exactly what Salin-Pascua l et al., (1992) found in a longterm study on sleep state misperception, that is, th e researchers were able to find those extreme sleep misperceives. They found that the sleep of misperceives followed a stable healthy sleep pattern, while they consistently complained of poor sleep. Single nigh t and averaging studies load the deck, as it were, to find over and underestimating sleep as a sleeper type and ignores the large amount of intraindividu al variability that may be occurring across days. Insomnia patients particularly have been shown to have hi ghly variable sleep patter ns (Coates et al., 1981; Edinger et al., 1997a, 1997b; Edinger et al., 1991 ; Frankel et al., 1976; Hauri & Wisbey, 1992; Mullaney, Kripke, & Messin, 1980; Vallieres, Ivers, Bastien, Beaulieu-Bonneau, & Morin, 2005; Vallieres & Morin, 2003) and is posited to be a central characteristic of insomnia (e.g., Espie, 2002; Frankel et al., 1976). Given that sleep misperception is common am ong insomnia patients and that insomnia patients are variable in both objective and s ubjective sleep parameters, the assumption that misperceptions are stable among insomnia patient s is preemptive and may not reflect the true patterns of sleep misperceptions. Assumptions th at sleep misperception reflects more negative
24 thinking and exaggeration may not reflect wh at is actually causing sleep misperception (e.g. Vallieres et al., 2005, p.452). Sleep researchers have acknowledged many problems with single night recording due to the high va riability in the sleep of insomn ia patients. The problems with averaging several nights to eliminate that va riability are less well recognized but gaining acceptance (e.g., Wohlgemuth et al., 1999). The re lationship between variability in sleep misperception and variability in sleep among insomn ia patients has not been tested longitudinally using multilevel modeling or intraindividual analys es. Such studies are needed before conclusion of sleep estimating sleeper types can be categorized based on singl e night and averaging techniques. Most people believe that averaging eliminates error. However, intraindividual variability represents true fluctuations in peoples sleep over time and is wa shed out by averaging. Like other behaviors that are highly variab le over time (Nesselroade, 2004; Nesselroade, Mitteness, & Thompson, 1984; Nesselroade & Salthouse, 2004), intraindividual variability analyses are essential for advancing the field of sleep. This study will employ both intraindividual analysis and multi-level modeling analysis techniques in this project. Sleep Misperception in Context Herein sleep misperception is defined as the difference between subjective and objective accounts of SOL and WASO. Unlik e other studies, this study does not assume that sleep misperception represents a percep tual distortion, nor a distinct sleep disorder. Indeed, sleep misperception may be more the researchers misp erception of what the pa tient perceives than vice versa. The term sleep misperception is ma intained in this study because the phenomena is presumed to be the same as that which other st udies have investigated. Regrettably though, sleep misperception has rarely been studied as an a ccurate perception of a unique physiological state that is neither wake nor sleep. Smith and Trinder aptly stated, [ O]verestimates of SOL may be a
25 normal response to poor sleep. If similar phenom ena occur in a clinical environment, the perception of disturbed sleep might be categorized as over estimation of SOL, when in fact the conventional sleep scoring underestimates the degree of the sleep disturbance (Smith & Trinder, 2000, p134). Given the crude measuring capability of sleep recording devices, such as actigraphy and EEG, the possibility remains that sleep mi sperception represents a sleep disturbance not detectible by current sleep measuring techniques. By comparing the daily phenomenon of sleep misperception to other daily sleep variables, such as total sleep time, the underlying mechanisms of sleep misperception may be better understood. Mechanisms of Sleep Misperception To date, research on the biopsychophysio logical m echanisms involved in sleep misperception is limited to a small number of EEG, actigraphy, and sleep diary studies. More broadly, biophysiological mechanisms of inso mnia are under studied and poorly understood (Drummond et al., 2004). There is substantial data to suggest that arousal is related to sleep misperception and may be a central mechanism of sleep misperception. Hyperarousal Many sleep experts agree that som e type of hyperarousal is involved in sleep misperception (Benoit & Aguirre, 1996; B onnet & Arand, 1997b; Drummond et al., 2004; Harvey, 2002a; Perlis, Giles, Bootzin et al., 1997; Pe rlis, Smith et al., 2001). At least one of three types of hyperarousal is e ndorsed by sleep researchers: cognitive (mental/emotional), physiological (somatic), and central nervous sy stem (CNS) hyperarousal. However, the line delineating these types of arous al is ambiguous and elusive. Some studies have shown a high correlation (Nicassio, Mendlowitz, Fussell, & Petr as, 1985) between cognitive and physiological arousal, while others suggest relative independe nce of these constructs (Craske & Craig, 1984). Many researchers hold that cognitive arousal is limited to the mind, while psychological arousal
26 refers specifically to activation of the sy mpathetic nervous system (Bonnet & Arand, 1997a; Nicassio et al., 1985; Schwartz, Davidson, & Goleman, 1978; Steptoe & Kearsley, 1990). CNS arousal involves greater cogniti ve activation (i.e., increased beta/gamma activ ity) during sleep. This activity may or may not be related to cogni tive or physiological ar ousal. (Drummond et al., 2004; Perlis, Merica et al., 2001; Perlis, Giles, Mendelson et al., 1997). The relationship of each type of hyperarousal with sleep mispercepti on is discussed in mo re detail below. Cognitive Arousal Conceptually, cognitive arousal is posited to be the degree to which the m ind is agitated and has been distinguished from physiological arousal and brain activa tion. Scientifically, like the mind, the construct cognitive arousal is difficu lt to isolate. Often, cognitive arousal is thought of as a top down process that can activate the st ress response but is not necessarily preceded by activation of the fight-or-flight response. Moreover, top down arousal, it is claimed, does not necessarily lead to activation of the sympathetic response. Seve ral studies have attempted to isolate the various types of cognitive arousal a nd their impact on sleep misperception. Intrusive thoughts (Lichstein & Rosenthal, 1980; Mitche ll, 1979; Monroe, 1967), sl eep threat monitoring, clock monitoring, rumination (Freedman & Sattler, 1982) and catastrophic worry have been linked to sleep misperception (Edinger, Wohlge muth, Radtke, Marsh, & Quillian, 2001; Harvey, 2002b; Harvey & Greenall, 2003; Lichstein & Rose nthal, 1980; Morin, Blais, & Savard, 2002). From this research, two types of cognitive arousal have been constructed: neutral arousal and anxious/emotional (Bonnet & Ar and, 1997a; Tang & Harvey, 2004). One study, in an attempt to contrast the impact of these two types of c ognitive arousal on sleep misperception in healthy sleepers, found that participants assigned to eith er anxious or neutral cognitive arousal conditions overestimated SOL, but only the cognitive arousal which involved anxiety, induced significantly lower estimates of TST than the control gr oup during a daytime nap (Tang & Harvey, 2004).
27 Physiological Arousal Tang and Harvey (2004) highlight the possibi lity that physiological arousal m ay cause sleep misperception and explain the high associa tion between sleep misp erception and insomnia. Individuals diagnosed with sleep state misperception show a variety of somatic markers of activation of the fight-or-flight response surrounding and durin g the sleep experience including high metabolic rate (Bonnet & Arand, 1997b). In addition, one study showed that brief global arousals lasting 3-30 seconds during sleep have been observed in patients with sleep state misperception (Smith & Trinder, 2000). This view is further strengthened by several studies which have scientifically manipulated sleep misp erception. In a series of studies, Mendelson and colleagues found that benzodiazepines (i.e ., Flurazepam, Zolpidem) reduced sleep misperceptions; more specifically, these drugs reduced the likelihood that insomnia patients reported EEG defined sleep as wakefulne ss (Mendelson, 1993, 1995a; Wallace B. Mendelson & Maczaj, 1990; Mendelson, Martin, Stephens, Giesen, & James, 1988). However, these drugs did not alter the perceptions of sl eep in normal sleepers (Mendelson, 1995b). It was concluded that benzodiazepines reduced arousal in these patients and simultaneously eliminated sleep misperception. Other studies have shown that caffeine administration induces sleep misperception, increased anxiety and, somatic hyperarousal in healthy sleepers (Bonnet & Arand, 1992, 1997a; Karacan et al., 1976; Okum a, Matsuoka, Matsue, & Toyomura, 1982). Researchers concluded that caffeine directly increases arousal which then leads to sleep misperception (Bonnet & Arand, 1997b; Tang & Harv ey, 2004). Collectively, these studies point to a strong relationship between physiol ogical arousal and sleep misperception. CNS Hyperarousal CNS hyperarousal (as determ ined traditionally by increased high frequency EEG in the beta range (14-35 Hz) and more recently by high metabolic rate in pos itron emission topography
28 (PET) and blood-oxygen level and decline (BOLD) in functional magnetic resonance imagining (fMRI)) is not only claimed to be the marker of sleep misperception (Perlis, Giles, Mendelson et al., 1997) but insomnia more generally (Drummond et al., 2004; Freedman, 1986; Merica, Blois, & Gaillard, 1998; Merica & Gail lard, 1992; Perlis, Smith et al ., 2001). The EEG measures brain wave activity by frequency and amplitude. High fr equency EEG activity in the beta [14-35 Hz] and gamma [35-45 Hz] range represents more brai n wave activity and dominates the active wake state of the brain. During healt hy sleep, EEG activity dramatically decreases and high frequency waves are replaced by progressively lower frequency higher amplitude activity in the .5-4 Hz range. Higher frequency activity in the beta/gamma range is characteristic of insomnia both at sleep onset and throughout the sleep period (reviewed in Perlis, Meri ca et al., 1998; Me rica et al., 2001). In addition, increased beta activity during th e lighter stages of sleep has been directly linked to sleep misperception (Perlis, Smith et al., 2001). Science has yet to show whether the length of time that beta activity persists into sl eep corresponds to insomnia patients experience of wakefulness. Sleep Misperception as Localized Sleep Deprivation Because all three typ es of arousals men tioned above are associated with sleep misperception and because the va lidity of delineating hyperarousal remains hotly contested (Drummond et al., 2004; Perlis, 2001), the current study will not parse out the source of arousal nor the causal relationship between arousal and sleep misperception. Sleep misperception may represent a true sleep disorder predicted by chan ges in the daily expression of sleep. Given the hypothesis that isolated areas of the brain remain active, or are prevented from sleeping deeply with the rest of the brain during sleep misperception, th ese areas may be subjected to a form of localized sleep deprivation. Global sleep deprivati on, for short and long periods of time, has been linked to reduced cognitive and physical performance, increased susceptibility to disease,
29 increased morbidity, reduced immune response, increased stress, and the onset of symptoms related to mental and physical health, such as depression, in somnia, and fibromyalgia. Like global sleep deprivation, localized sleep deprivation may predis pose individuals to problems related to the areas of the brain that are affected in sleep misperception. Importantly, sleep misperception may represent an acc urate perception of a sleep problem that is undetectable by standard EEG measures. Regardless, this pheno menon may have serious financial and health consequences. Sleep misperception may result from the areas of the brain involve d in consciousness or executive functioning remaining active, consequent of hyperarousal while the rest of the brain has entered sleep (Merica et al., 1998). This conc lusion is consistent with the localization theory of sleep (Krueger & Obal, 1993). They argu e that sleep is not a whole brain event and that each area of the brain has its own circadian rhythm and homeostatic need. Merica and colleagues (1998), applied the local sleep theory to slee p misperception. In their neuronal transition probability model (NTP), they argue that areas of the brain that remain active during EEG sleep may represent an intermediate or transiti on state between sleep and wake where sleep misperception occurs (Merica et al., 1998). The subtle difference in these perspectives is that the later continues to delineate wake state from i ntermediate state from sleep state. Localized theory recognizes that sleep and wake are no t discrete states but can co-occur by degree. Regardless of this distinction, bo th perspectives highlight the po ssibility that whole body arousal is not necessarily required for lo calized areas of the br ain to stay active du ring sleep. Only those areas of the brain involved in pe rception would need to be active to produce what is called sleep misperception. Although beta activity has been gl obally identified in conjunction with sleep misperception, the location of beta activity has no t been determined. Howe ver, beta activity is
30 most likely occurring in only isolated areas of th e brain related to attention, monitoring and selfawareness, such as the dorsal la teral prefrontal cortex, the anteri or cingulate gyrus, and insula. In this study, distinct areas of the brain involved in sleep mi sperception were not directly investigated. Nonetheless, the understanding of sleep misperception as a localized brain event provided the theoretical foundation on which the anal ysis and interpretation of results were built. Because objective changes in the sleep of patie nts with sleep misperceive is lacking, many have argued that misperception is an arousal problem not relate d to a sleep disturbance. For example, Bonnet and Arand (1995) argue that slee p misperception is related to increased arousal among individuals with longer sleep need. Howeve r, sleep requirement/length is determined by two sleep systems rather than one conceptua lized in Bonnets model. The preeminent sleep theorist and researcher Alexander Borbly (1982 ) posited two major brain processes essential to sleep. His theory, called the tw o process model of sleep, has dominated basic sleep research for the past 2 decades and continues to influe nce the field. The first of his processes, the circadian process, or more simply the C-process is what most people think of when they think about sleep. The C-process is the bi ological drive to sleep that is unique to each species and to a lesser extent to each individual wi thin each species. This process is what makes humans diurnal and bats nocturnal. The C-process is driven by bi ological clocks in the brain, which rhythmically increase and decrease the brains propensity to sleep on about a 25-hour basis. The other process in Borblys model is called the homeostatic or S-process. This process increases the brains propensity to sleep as a function of time awake a nd the intensity at whic h neurons are used. The idea here is that the more an area of the brain is used the more that area needs sleep. Following 24-hour sleep deprivation, the S-pro cess intensifies the dr ive to sleep late in the night and into
31 the next day while the C-process actually begins to decrease its in fluence on sleep. These processes have been tied to insomnia and ar e applied to sleep misperception in this study. Sleep misperception may be related to a homeostatic dysregulation. Following sleep deprivation, sleep rebounds by exte nding the amount of sleep. This is because it takes longer to restore sleep homeostasis followi ng deprivation. The circadian pro cess is not altered by sleep deprivation but is highest roughl y 16 hours after waking. Tenuous evid ence exists that some type of C-process dysregulation is involved in inso mnia (reviewed in Pige on & Perlis, 2006). For example, abnormal temperature rhythms have been observed in insomnia (e.g., Morris, Lack, & Dawson, 1990). However, these changes may be the result of protective behaviors often observed in insomnia patients who alter the timi ng of light exposure and biological rhythms, such as temperature (Pigeon & Perlis, 2006). A nother study found that in relationship to sleep misperception, circadian dysregulation is most li kely a consequence of sleep misperception rather than a causal factor. Nonetheless, circadia n rhythms must be consid ered in relationship to sleep misperception because before sleep mispercepti on could occur most of the brain must be in a sleep state. The C-process is the major force th at drives the global sleep state. Sleep propensity imposed on the brain by chronobiolog ical clocks that re gulate sleep must be highly active in the presence of high arousal in percep tual centers of the brain. When arousal is relatively higher than the S-process of a given area of the brain, and the C-process is higher than both the arousal system and S-process, then global sleep will be observed but localized sleep deprivation will occur in the isolated areas. Local sleep th eory would predict that if arousal in the perceptual/consciousness centers of the brain exceeds the global sleep drive (the C-process) and also the local sleep drive (S-process), that area may remain wake; thus, perceiving a wake state and the passing of time. If sleep were restricted, the area of the brain that remained awake during
32 global sleep would receive, potenti al, much less sleep than the rest of the brain. For example, if sleep were restricted to 6 hours, but the perceptual areas of th e brain remained active for 1 hour, the patient would report that they had only received 5 hours in c ontrast to the objective 6 hours that was observed using EEG. However, it may be possible that local sleep loss, such as is posited here, may lead to sleep rebound in that ar ea and on the whole brain. That is, if sleep were not restricted the patient may aw ake less during the night (or perc eive that they had awoken less during the night) and sleep longer in the morning. In fact, this model would predict that even if the global brain awoke, the patient might not be aw are because the perceptual areas of the brain would still be sleeping.. Aims 3 and 4 i nvestigate the relationship between SOLsm, WASOsm and TSTo from this perspective. Summary In summ ary, sleep misperception is operationa lly defined in this st udy as the discrepancy between objectively measured sl eep and the participants per ceptions. The idea that sleep misperception represents a perceptual deficit or exaggeration of sleep difficulty by a select group of sleep complainers, though ubiquitously accepte d, has little scientific support. That sleep misperception increases with age is well known; however, much research is needed to better understand the patterns and corre lates of sleep misperception in this population. Arousal is strongly associated with sleep misperception. Th is study is based on the theoretical assumption that sleep misperception may be a form of lo calized sleep deprivation resulting in sleep deprivation-like consequences (e.g., longer total sleep time).
33 CHAPTER 3 STATEMENT OF THE PROBLEM The m ain objective of this study is to better understand sleep misperception (SM) by investigating the distribution, patte rns and correlates of subjective relative to objective measures of SOL and WASO, hereafter called SOLsm and WASOsm, in a sample of community dwelling older adults. Daily and biweekly averages of SOLsm and WASOsm will be studied in relation to1) the amount of withinto between-person variability, 2) sleep complaint status, 3) each other, and 4) actigraphy measured total sleep time (TSTobjective). Specific Aim 1 The f irst specific aim is to investigate SOLsm and WASOsm across two weeks. Aim 1 will determine the amount of withinto between-perso n variability in SOLsm and WASOsm (as determined by daily diary self-report of SOL or WASO minus SOL or WA SO as determined by daily actigraphy measures, respectively). Hypothesis for Specific Aim 1 It is hypothesized that there wi ll be greater intraindividual variability (i.e., within-person variab ility) than between-p ersons. It has been reported that people can be categorized as either positive or negative perceivers of their sleep (Means et al., 2003; Trajanovic, Radivojevic, Kaushansky, & Shapiro, 2007). However, this assert ion is based on research that looks only at a single night of sleep or averages over several nights of sleep. These appr oaches ignore the potential confound that individuals SOLsm or WASOsm may be highly variab le across several nights. As a result these studies may only capture individuals on the extreme of a continuum of misperceiving sleep. Intraindiv idual variability in SOLsm and WASOsm have not been tested to confirm or contradict the asserti on that there are sleep perceiver t ypes that consitantly either over or under perceive SOL and WASO.
34 Specific Aim 2 The second specific aim is to investigate SOLsm and WASOsm among older participants who reported a sleep problem (complainers) and those that did not ( non-complainers). Aim 1 will determine the amount of within-t o between-person variability in SOLsm and WASOsm that occurred among sleep complainers as compared to non-complainers. Hypothesis for Specific Aim 2. Patients with insomnia experience high night-to-night variability in sleep. High intraindividual variab ility among insomnia patients has been reported for both objective and subjective meas ures of sleep onset latency (SOL), total sleep time (TST), waketime after sleep onset (WASO), and subjective sleep quality (SSQ) (Edinger, Marsh et al., 1991). In spite of this high degree of within-patie nt variability across a myriad of sleep measures, it has been reported that insomnia patients consistently misperceive the sleep experience, including overestimating SOL and WASO. Within-person variability in SOLsm and WASOsm among individuals who complain of sleep difficul ties has not been compared to individuals who do not complain of sleep difficulties. The analys es preformed in aim 2 should help determine whether sleep misperception is a co nsistent trait-like behavior of a subgroup of individuals or a behavior that manifests by degree among older adults. Previously our lab showed that sleep complain t among older adults rela ted to an occasional pattern of extreme sleep overestimating SOL and WASO that were more extreme but not more frequent among individuals with sleep complaint than those without complaint (Kay, McCrae, & Rowe, 2008). In that study, a one-way MANOVA revealed a significant gr oup effect for the two sleep misperception dependent variables (F4,89 = 7.65, p < .001). Follow-up testing revealed complainers over-estimated WASO more frequently than non-complainers ( Mcomplainers = 3.67 0.74 days versus Mnon-complainers = 1.27 0.22 days; F1,92 = 14.85, p < .001). Frequency of SOL over-estimation did not differ by group(~7 days; F1,92 = 0.01, p = .94). Interestingly, on nights
35 that over-estimation occurred, th e average amount over-estimated was greater for complainers than non-complainers for both SOL ( Mcomplainers = 22.54 2.74; Mnon-complainers = 14.21 1.17; F1,92 = 9.63, p < .01) and WASO ( Mcomplainers = 44.62 10.01; Mnon-complainers = 14.48 3.54; F1,89 = 13.13, p < .001). For both groups, SOL over-estim ates did not correlate with WASO overestimates. Importantly, over-estimating did not occur daily among complainers; however, on days over-estimating occurred, their estimates were more extreme on average than noncomplainers. Intraindividual analyses which compare the amount of withinto between-person variability by complaint status can be used to he lp determine if SM is a between-persons trait or a highly variable pattern w ithin-persons. Aim 2 will test the hypothesis that SOLsm and WASOsm will be highly variable within individuals regardless of sleep complaint status. Specific Aim 3 To investiga te the relationship betw een day-to-day fluctuations in SOLsm to fluctuations in WASOsm in older adults. Hypothesis for Specific Aim 3. Most studies of sleep misperception have focused on SOL. Perlis et al. (1997) argued that because high frequency EEG activity is central to both sleep onset and sleep misperception, SM should be mo re prominent during that period. Moreover, Mendelson (1998) found that sleep perceptions wher e more congruent with PSG during the third sleep cycle as compared to cycles 1 and 2 am ong insomnia patients. However, SM can and does occur during WASO also. Sleep maintenance proble ms that contribute to WASO event are much rarer in younger adults than ol der adults, even among insomnia populations (Bliwise, 2005). In late life, nighttime awakenings become more fr equent and present greater opportunity to study sleep misperception than occur s pontaneously during the night. Tr aditionally, sleep researchers have combined SOLsm and WASOsm across days to create a more robust SM variable (Means et
36 al., 2003). Because combining multiple nights of SOLsm and WASOsm into one variable creates a more robust variable, it is likely that these variables co-occur on some level. However, it remains unclear whether day-to-day fluctuations in SOLsm and WASOsm vary together or fluctuate independently from one another. That is, on days that WASOsm are high does that predict that SOLsm will also be high. The positive relationship between SOLsm and WASOsm has only been demonstrated on the average but not the daily level. For example, one study that investigated SM in insomnia patients showed that participants perceived sleep as wakefulness during all lighter stages of sleep (i.e., stages 1 and 2) thr oughout the night (Mercer, Bootzin, & Lack, 2002). However, the averaging techniques used in that study prevented the researchers from concluding that those who misperceived SOL were the same individuals who misperceived WASO. Ultimately, they showed only that both SOLsm and WASOsm are common among insomnia patients. The question remains, are percepti ons of SOL and WASO independent from or dependent on each other. Studies that investig ate the night-to-night relationship between SOLsm and WASOsm are needed to determine th eir longitudinal relationship. Many people believe that sleep misperception re presents a negative bias imposed on sleep or an exaggeration of sleep difficulties. This a ssumption would predict that a negative bias would be imposed consistently on all aspects of sl eep. While the consistency in average sleep misperception is well established, consistency as it relates to night-to-night consistency and variable-to-variable (e.g., SOLsm to WASOsm) consistency has not been tested. This assumption would be greatly weakened if it can be shown that there is no relationship between SOLsm and WASOsm on a night-to-night basis. It is hypothesized that WASOsm and SOLsm will not vary together night-to-night. More specifically, it is h ypothesized that, even though on an average
37 SOLsm and WASOsm are positively re lated, greater SOLsm will not necessitate greater WASOsm and vice versa on a daily level. Specific Aim 4 To determ ine the intraindividual daily relationship between SOLsm and WASOsm to objectively measured TSTo (actigraphy). Hypothesis for Specific Aim 4. It is hypothesized that with in-person fluctuation in SOLsm and WASOsm will share a systematic relationship to fluctuation in TSTo. Bonnets model of sleep posits that SM is the interaction between the slee p and arousal systems, such that longer sleep requirement coupled with chronic hyperarousal is the formula for SM. Indeed, insomnia patients who sleep longer than others are more likely to be identified as having sleep misperception. Moreover, our lab has previously shown that older adults who complained of hyperarousal related health conditions, such as chronic pain and depression, have greater SM than noncomplainers regardless of whether there was a sleep complaint or not (Kay, McCrae, & Rowe, 2008). Bonnet and Arand (1995) assumed that sleep re quirement is a trait-lik e characteristic of sleepers. However, sleep requirement may vary significantly day-to-day. Increased TST is a marker of an S-process sleep dysregulation. Sleep deprivation is the most salient way to disrupt the S-process. During sleep deprivation, sleep pr omoting substances accumulate in the brain and exert their effects once sleep is obtained. This disruption to the S-Process through sleep deprivation can induce increased TST. We posit that SM represen ts a form of localized sleep deprivation to perceptual areas of the brain. While global sleep is obtained, the areas responsible for conscious awareness remain active and are, thus, prevented from obtaining the localized sleep need. As a result, SM as a localized sleep deprivation may predict a similar sleep rebound in TST as global sleep deprivation. Base d on this theory, we predict that TSTo will increase with increased SOLsm and WASOsm from night-to-night. Aim 4 will test the hypothesis that, within-
38 persons, on days that individuals have greater SOLsm or WASOsm than usual, his/her TSTo will be longer than usual.
39 CHAPTER 4 METHODS Participants This study involved secondary analyses on a database com piled by McCrae and Rowe (2003) in a study on sleep in older adults. A convenience sample of 116 community-dwelling older adults ( Mean age=72.81, SD =7.12) were recruited from the North Florida area. A variety of recruitment techniques were employed incl uding media advertisem ents, community groups, and flyers. Recruitment materials described the research as a study of sleep patterns in the elderly. Participants were compensated $30 for th eir participati on. Interested individuals were screened in two phases to determine if they met the criteria for inclusion. Phase one consisted of a brief telephone interview (15-20 minutes), a nd phase two involved an in-person interview either in the individuals home (76%) or at a local continuing care retirement center (24%). Individuals were excluded on th e basis of six exclusionary criteria: 1) age younger than 60 years; 2) self-report of sleep disorder diagnoses other than insomnia (e.g., sleep apnea or narcolepsy); 3) self-report of sleep symptoms i ndicative of sleep diagnoses other than insomnia (e.g., heavy snoring, gasping for brea th, leg jerks, daytime sleep at tacks); 4) presence of severe psychiatric disorders (e.g., thought disorders or depression); 5) cognitive impairment, scoring in the impaired range on three or more subtes ts of the Neurobehavioral Cognitive Status Examination (Cognistat; Mueller, Kiernan, & Lang stron, 2001); 6) use of psychotropic or other medications know to alter sleep (e.g., beta-block ers); and 7) medical c onditions that impaired independent daily functioning (McCrae et al., 20 05). Participants who en tered the study were gave informed consent in accordance with the sta ndards of the University of Florida Institutional Review Board.
40 Of the 116 individuals recruited, 103 were enro lled in the study. Thirteen individuals were ineligible to participate in the study due to ag e, dementia, medication, an d sleep apnea diagnosis. The mean age of the participants was 72.81 years ( SD = 7.12). The majority of participants were European Caucasian (96%), female (66%), college educated (75%; M = 16.34 years, SD = 2.92), and married (59%). All of the participants lived in their own homes during the study. Procedures Data was collected at three periods during the st udy: baseline, end of first week, and end of second week. During the initial 11 hour interview, participants read and signed an inform ed consent form approved by the University of Fl orida Institutional Review Board. Once consent was obtained, the Cognistat, and the demographics and health survey were administered by a member of the research team At this time, both the sleep diaries and the Actiwatch-L (ACT-L; Mini Mitter, Inc.) were explained to the participants. The part icipants were advised to complete the sleep diaries and wear the Actiwatch-L continuously for 14 days. At the end of the first week, the sleep diaries were collected from the participants and the da ta was downloaded from the Actiwatch-L. At the end of the second week, the final week of sleep diaries and ActiwatchL data were collected. The Beck Depression Inventory-Second Edition (BDI-II Beck, Steer, & Garbin, 1996) was also completed at this time. Participants wore the Actiwatch-L wrist device both day and night for two weeks to determine objective total sleep time (TSTo), sleep onset latency (SOLo) and wake-time after sleep onset (WASOo). While wearing the Actiwatch-L, participants also completed daily sleep diaries, recording their subjective es timates of, sleep onset latency (SOLs) and wake time after sleep onset (WASOs).
41 Measures Overviews of the sleep variables, the demogr aphics and health survey, the Cognistat, and the measures of daytime func tioning are presented below. Objective Sleep Variables Objectiv e sleep was measured using the Actiwatch-L(Mini Mitter Co., 2001). Within the Actiwatch-L, data is sampled 32 times per second over a 30 second epoch using an omnidirectional, piezoelectric accel erometer with a sensitivity of > 0.01 g-force. A sum of the peak activity counts for each 30 second epoch is downloaded to a PC and then analyzed by Actiware-Sleep vol. 3.3. (Mini Mitte r Co. Inc., 2001). Three sensitiv ity settings (high, medium, and low) are provided by the software for detect ing wake/sleep periods. A high sensitivity setting was used in the current study since it provides high correlations with PSG measured total sleep time (.95) for healthy older adults (Colling et al., 2000) and for total sleep time (.73) and sleep onset latency (.93) for individuals with insomnia (Cook et al., 2004). Additionally, actigraphy has valid criterion-validity when compared to PSG (.80) and high test-retest reliability (0.92; (Ancoli-Israel et al., 2003; de Souza et al., 2003; Hauri & Wi sbey, 1992). A validated algorithm is used to identify the activit y of each epoch as wake or sleep (Oakley, 1997). With the high sensitivity setting, the threshold for wake is 20 activity counts. If the peak activity count for an epoch is > 20, the epoch will be scored as wake. If th e peak activity count for an epoch is < 20, the wake/sleep determination is made based on th e activity that occurs in the two-minute period surrounding the epoch. The wake/sleep determination if the activity count is < 20 is made based on following equation: Total Activity for Epoch A = EA-4 (.04) + EA-3 (.04) + EA-2 (.20) + EA-1 (.20) + E (2) + EA+1 (.20) + EA+2 (.20) + EA+3 (.04) + EA+4 (.04)
42 where A = # of activity counts for the epoch being scored; EA +/1-4 = # of activity counts in adjacent epochs. If the Total Activity for Epoc h A (weighted sum of activity counts) exceeded the threshold value of 20, then Epoch A is scored as wake; otherwise, it is scored as sleep (McCrae et al., 2005). Using the Ac tiware-Sleep vol. 3.3 software (Mini Mitter Co. Inc., 2001), a number of sleep parameters are derived from th e data including sleep ons et latency, total sleep time, and wake after sleep onset. The definitions of the objective sleep variables used in this study are: sleep onset latencyo (interval between bedtime and sleep start); total sleep timeo (time in bed minus SOLo, WASOo, and time spent in bed between awakening in the morning and getting out of bed); and wake time after sleep onseto (time spent awake afte r initial sleep onset until last awakening). The subscripts s and o are used to distinguish between the subjective and objective sleep variables. Actigraphy is less reliable at determining WASO than it is at identi fying sleep onset or TST. Compared to PSG, actigraphy systematically underestimated WASO (Cole, Kripke, Gruen, Mullaney, & Gillin, 1992). This is explained by th e relatively lower resolution of actigraphy at detecting wake compared to detecting sleep. In field studies, however, actigraphy remains a good measure of detecting WASO in comparison to PSG, particularly when the WASO event lasts longer than 15 seconds. Actigraphy detects 88% of such events (Horne, Pankhurst, Reyner, Hume, & Diamond, 1994). Subjective Sleep Variables Sleep diary. Subjective s leep quantity was measured using sleep diaries. The sleep diary (Lichstein, Riedel, & Means, 1999) was completed by each participant each morning for 14 days (see Appendix A). Two subjective sleep measures were obtained from the sleep diary data: sleep onset latency (SOLs; initial time from lights out until sleep onset), and wake time after sleep onset (WASOs; time spent awake after initial sleep onset until last awakening). Sleep diaries are
43 commonly used in research, have been validat ed in several studies, and are accepted as a scientifically useful measures of the subjec tive sleep experience (Espie, Inglis, Tessier, & Harvey, 2001). Sleep Misperception Variables (SOLsm and WASOsm) In this study, the difference between objective and subjectiv e measures are called sleep misperception. Raw daily subjective reports of SO L and WASO were compared to respective daily actigraphy measures to compute daily SOLsm and WASOsm (subscript sm denotes sleep misperception). The equation for each variable is listed below: SOLsm = SOLs SOLo WASOsm = WASOs WASOo Daytime Functioning Measure Beck Depression Inventory Second Edition (BDI-II) Depression was m easured using the Beck Depression Inventory-Second Edition (BDI-II; Beck, Steer, & Garbin, 1996). This is a 21-it em measure with a scale ranging from 0-3 measuring the severity of depressive symptoms (3 being the most severe). Scores range from 0 63. Scores within the 0 13 ra nge indicate none or minimal depression, 14 to 19 indicate mild depression, 20 to 28 indicate moderate depressi on, and 29 to 63 indicate severe depression. Participants were asked to respond to the ques tions based on the previous two weeks to match the two-week sleep and actigraphy recording pe riod. The BDI-II has demonstrated sufficient internal consistency reliability (.90) and concurrent validity ( .69 .76; Storch, Roberti, & Roth, 2004). Demographics and Health Survey This survey consists of 13 item s collecting information on demographics, sleep disorder symptoms, physical health, and mental health (s ee Appendix A; Lichstei n et al., 2004). Health
44 conditions were assessed as the number of condi tions selected from the following list: heart attack, other heart problems, cancer, AIDS, hype rtension, neurological di sorder (e.g., seizures, Parkinsons), breathing disorder (e.g., asthma emphysema, allergies), urinary problems (e.g., kidney disease, prostate problem s), diabetes, pain (e.g., arthriti s, back pain, migraines), and gastrointestinal disorders (e.g., st omach, irritable bowels, ulcers, ga stric reflux). Self-report sleep questions on the survey contained information on whether the participan t had a sleep problem and if they or a bed partner noticed heavy s noring, difficulty breathing or gasping for breath, frequent leg jerks, restlessness before sleep onset, sleep attacks during th e day, or paralysis at sleep onset. Two health survey variables were used in this study; sleep complaint and the total number of health complaints. Sleep Complaints: Determined by yes/no repo rt to the question: Do you have a sleep problem? yes or no Number of Health Cond itions: The number of health conditions that the participants complained of, except sleep and pain) were summ ed. Scores could range from 0-10 complaints. Data Analyses The current study applied several statistical techniques to investigate the four specific aim s of this study including intraindividual analys es and multilevel modeling (MLM). The analytical techniques used for each aim are explained and described below. Specific Aim 1 The f irst specific aim is to investigate SOLsm and WASOsm across two weeks. Aim 1 will determine the amount of within-to between-person vari ability in SOLsm and WASOsm. Data analysis. Intraindividual variability (IIV) analyses will be used to determine the relative amount if withinto be tween-person variability in SOLsm and WASOsm. To control for potential systemic growth in the data, all variables will be de-trended prior to calculation of
45 indexes of withinand between-person variability. To de-trend the data, linear regression will be run with SOLsm and WASOsm as the independent variables. The subsequent unstandardized residual values resulting from th e regressions will then be saved and used as time independent values. Using the residual values computed a bove, an index of between-person variability (Sample Standard Deviation, SD) and within-person vari ability (Individual Standard Deviation, ISD) will be computed. These values will then be compared by dividing the ISD by the SD to get the proportion of between-person variab ility that is f ound within-persons. Specific Aim 2 The second specific aim is to compare the amount of withinto betw een-person variability in SOLsm and WASOsm among sleep complainers with that among sleep noncomplainers. Data analysis. Intraindividual variability analyses will be used to determine the relative amount if withinto between -person variability in SOLsm and WASOsm within each of the sleep groups (i.e., sleep complaining and sleep non-co mplaining). The de-trended data obtained following aim one will be separated into the respective two groups. Two separate linear regressions (one for each gr oup) will be run with SOLsm and WASOsm as the independent variables. The subsequent unstandardized residual values resulting from the regressions will then be saved and used as time independent values. Using the residual values computed above, an index of between-person variability (Sample Standard Deviation, SD ) and within-person variability (Individual Standard Deviation, ISD) will be computed for each group. For each group of data, these values will then be compar ed by dividing the ISD by the SD to get the proportion of between-person variability that is found within-persons. Specific Aim 3 To investiga te the relationship betw een day-to-day fluctuations in SOLsm to fluctuations in WASOsm.
46 Data analysis. A multilevel modeling (MLM) approach was used to predict within-person and between-person relationships between SOLsm and WASOsm. MLM, also referred to as hierchical linear modeling (HLM; Bryk & Rauenbush, 1992) is an extension of the general linear model and is especially suited for daily data, such as sleep, because of their autoregressive nature and hierarchical structure with daily observations nested with in each participant (Singer, Davidson, Graham, & Davidson, 1998; Singer, Fulle r, Keiley, & Wolf, 1998; Singer & Willett, 2003). Because of the hierarchical na ture of the data used in this study (14 days consecutive days nested within 103 participants) and in order to in crease the precision of predicting fluctuations in SOLsm with fluctuation in WASOsm), the data will be modeled with a MLM approach. This provides the opportunity to examine the relations hip between these vari ables on two levels: a within-(level 1) and a between (level 2)-pers ons. Level 1 analysis addresses two questions: First,On days in which a person reports above-average SOLsm, does s/he also report aboveaverage WASOsm? and second, On days in which a person reports above-average WASOsm, does s/he also report above-average SOLsm? This level of analysis is concerned with questions of atypical days within an indi vidual and what variables fluc tuate, within-persons, on these atypical days. Level 2 analyses examine the question: Do people who generally have high WASOsm also report high SOLsm and vica versa? WASOsm will be used to predict SOLsm and vice versa in two separate models, each using a three step MLM approach. In the first model, Step 1 (Tables 5-3, Row 1) the null (baseline) model, will estimate only a fixed and random intercept for WASOsm(Bryk & Raudenbush, 1992). This model will specify that WASOsm for person j is a function of the overall group-average WASOsm, a between-person random error term, and a within-person random residual component. This step provides a comparison for the step 3 model, which includes SOLsm. In step 2, time functions (linear) will be
47 added as a covariate to the null model (Table 5-3, Row 2), producing a latent growth curve model. As such, the model will specify that WASOsm for person j on day i is a function of: average WASOsm, linear time, a between-persons random error term, and a within-person random residual component. Next, SOLsm will be added to the model. In step 3 (Table 5-3, Row 3) the estimates of the fixed and random intercep ts and fixed linear slopes for each sleep variable will be added. Thus, the daily WASOsm for each person will be predicted by: average level of WASOsm, linear time, the between-person effects of mean-level SOLsm, the within-person effects of SOLsm, a between-person random error term, and a within-person random residual component. In the second model, Step 1 (Tables 5-5, Row 1), the null (baseline) model, will estimate only a fixed and random intercept for SOLsm (Bryk & Raudenbush, 1992). This model will specify that SOLsm for person j is a function of the overall group-average SOLsm, a betweenperson random error term, and a within-person random residual component. This step provides a comparison for the step 3 model, which includes WASOsm. In step 2, time functions (linear) will be added as a covariate to the null model (Tab le 5-5, Row 2), producing a latent growth curve model. As such, the model will specify that SOLsm for person j on day i is a function of: average SOLsm, linear time, a between-persons random erro r term, and a within-p erson random residual component. Next, WASOsm will be added to the model. In step 3 (Table 5-5, Row 3) the estimates of the fixed and random intercepts and fixed linear slopes for eac h sleep variable will be added. Thus, the daily SOLsm for each person will be predicted by: average level of SOLsm, linear time, the between-person effects of mean-level WASOsm, the within-person effects of WASOsm, a between-person random erro r term, and a within-person random residual component. All models will be estimated using the Maximum Likelihood (ML) method. The ability of the first model to predict WASOsm and the second model to predict SOLsm better than the
48 baseline model (i.e., Deviance) will be used as an index of Goodness of Fit. Improvements in predictability will be determined by the amount of reduction of withinand between-person residual variances compared to the baseline model (Bryk & Raudenbush, 1992). Decreases in residual and intercept variances represent a proportional reduction of the prediction error, which is analogous to R2, and will be used as an estimate of withinand between-person effect sizes. The amount of agreement between model predicted va lues and actual values will be calculated as an estimate of an overall effect size. Specific Aim 4 To determ ine the intraindividual daily relationship between SOLsm and WASOsm to TSTo. Data analysis. A multilevel modeling (MLM) approach was used to predict within-person and between-person relationships of SOLsm and WASOsm to TSTo. WASOe and SOLsm will be used to predict TSTo using a four step MLM approach. Step 1 (Table 5-7, Row 1), the null (baseline) model, will estimate only a fixed and random intercept for SOLsm (Bryk & Raudenbush, 1992). This model will specify that TSTo for person j is a function of the overall group-average TSTo, a between-person random error term, and a within-person random residual component. This step provides a comparison for the step two models which includes time. In step 2, time functions (linear) will be added as a covariate to the null model (Table 5-7, Row 2), producing a latent growth curve model. As such, the model will specify that TSTo for person j on day i is a function of: average TSTo, linear time, a between-pers ons random error term, and a within-person random residual component. In step 3 (Table 5-7, Row 3) the estimates of the fixed and random intercepts and fixed linear slopes for each sleep variable will be added. Thus, the daily TSTo for each person will be predic ted by: average level of WASOsm, linear time, the between-person effects of mean-level WASOsm, the within-person effects of WASOsm, a between-person random error term and a within-person random residual component. In step 4
49 (Table 5-7, Row 4) the estimates of the fixed an d random intercepts and fixed linear slopes for SOLsm will be added. Thus, the TSTo for each person will be pred icted by: average level of TSTo, linear time, the between-perso n effects of mean-level SOLsm aned WASOsm, the withinperson effects of daily-centered SOLsm and WASOsm, a between-person random error term, and a within-person random residual component. All models will be estimated using the Maximum Likelihood (ML) method. The ability of a model to predict TSTo better than the baseline model (i.e., Deviance) will be used as an index of Goodness of Fit. Improvements in predictability will be determined by the amount of reduction of withinand between-p erson residual variances comp ared to the baseline model (Bryk & Raudenbush, 1992). Decreases in residu al and intercept va riances represent a proportional reduction of the predicti on error, which is analogous to R2, and will be used as an estimates of within and between-person effect sizes. The amount of agreement between model predicted values and actual values will be calculated as an estimate of an overall effect size.
50 CHAPTER 5 RESULTS The m ain objective of the study was to better understand sleep misperception by investigating SOLsm and WASOsm in a sample of community dwelling older adults. This objective was achieved by examining daily values and biweekly averages of SOLsm and WASOsm in relation to1) the amount of withinto between-person variability, 2) the amount of withinto between-person variab ility by sleep complaint status, 3) each other, and 4) objective total sleep time (TSTo). The main objective was achieve d through these four aforementioned specific aims. Results will be discussed separately for each aim. Specific Aim 1 The f irst specific aim was to investigate SOLsm and WASOsm across two weeks to determine the amount of withinto between-perso n variability in SOLsm and WASOsm. Intraindividual variability (IIV) analyses were us ed to determine the relative amount of withinto between-person variability in SOLsm and WASOsm. Dissection of SOLsm and WASOsm variables revealed a cons iderable amount of withinto betwee n-person variability suggesting that individuals are highly variable in sleep misperception across two weeks. There was greater within-person variance in participants SOLsm than between-person variance, and almost as much within-person variance in participants WASOsm as between-persons. Specifically, SOLsm was found to vary within-persons over 150% as mu ch as it varies between-persons and withinpersons WASOsm displayed 95% of the amount of between -person variability. For a listing of the amount of within-person variability compared to between-per son variability, see Table 5-1. For a graphical depiction of the rela tive amount of within-person to between-person variability, see Figure 5-1.
51 0 0.5 1 1.5 2 SOL WASOAmount of withinto betweenperson variability (% ile) Figure 5-1. Relative amount of w ithin-person variability compared to between -person variability after controlling for any linear or quadratic effects of time. Specific Aim 2 The second specific aim was to determine the amount of withinto between-person variability in SOLsm and WASOsm by sleep complaint status. Intraindividual variability analyses were used to determine the relative amount if withinto between-person variability in SOLsm and WASOsm within each of the sleep groups (i.e., sleep complaining and sleep non-complaining). As compared to other individuals with the same complaint status, participants were nearly as, and generally more variable within-persons as compared to between-persons in SOLsm and WASOsm. Both the sleep non-complaining and compla ining groups had large amounts of withinperson variability in their sleep estimates as compared to between-person variance in sleep estimates. However, there was great variability in the complainers in both SM variables (Figure 5-2).
52 Figure 5-2. Relative amount of w ithin-person variability compared to between -person variability after controlling for any linear or quadratic effects of time. Specific Aim 3 Aim 3 was to investigate the relationship between day-to-day fl uctuations in SOLsm and WASOsm. A multilevel modeling (MLM) approach was used to predict within-person and between-person relatio nships between SOLsm and WASOsm. Because the self-reported perception of SOL and WASO were recorded by the particip ants at the same time each day, and because the experience of either could impact the percepti on of the other while re porting, determining the relationship of SOLsm and WASOsm requires that two models be built: one using SOLsm to predict WASOsm and the other us ing WASOsm to predict SOLsm. Multilevel Model for WASOsm The intraclass correlation coefficient (ICC), which serves as an index of the amount of withinand between-person va riability to be explained (Bryk & Raudenbush, 1992), was 0.41. This indicates that, 41% of the overall variability in WASOsm was within-person and 59% was between-person. For a complete listing of model parameters and estimates obtained at each step of the model building process, see Table 5-3. 0 1 2 SOLWASOAmount of withinto between-p e variability (%-ile) Non-complainers Complainers
53 In the final MLM for WASOsm, SOLsm was a significant between-person, level 2, predictor, = 0.62, t (101.43) = 3.29, p < .01. At the within-person leve l, level 1, the predictor Day, = -0.03, t (129.50) = -3.11, p < .05, was significant. There was a nonsignificant trend at the within-person level, le vel 1, for the predictor SOLsm, = -0.10, t (91.09) = -1.78, p = .08. The model also contained a signifi cant random effect of SOLsm, = 0.13, Walds Z = 3.83, p < .01. This model explained approximately 10% of the within-person variance and 20% of the between-person variance. See Table 5-4 for a total listing of predictor estimates and significance levels. Multilevel Model for SOLsm The intraclass correlation coefficient (ICC), which serves as an index of the amount of withinand between-person va riability to be explained (Bryk & Raudenbush, 1992), was 0.17. This indicates that, 83% of the overall variability in SOLsm was within-person and 17% was between-person. For a complete listing of model parameters and estimates obtained at each step of the model building process, see Table 5-5. In the final MLM for SOLsm, WASOsm was significant between-person, level 2, predictor, = 0.21, t (98.09) = 3.98, p = .01. At the within-person level, level 1, the predictors Day, = 0.74, t (204.24) = -3.47, p < .01, and WASOsm, = -0.1, t (87.59) = -2.01, p < .05, were significant. The model also containe d a significant random effect of WASOsm, = 0.13, Walds Z = 4.02, p = .01. This model explained approximately 8% of the within-person variance and 17% of the between-person variance See Table 5-6 for a total listing of predictor estimates and significance levels.
54 Specific Aim 4 Ai m 4 was to determine the intraindividua l daily relationship between SOLsm and WASOsm to TSTo. Multilevel modeling (MLM) was used to predict within-person and between-person relationships of SOLsm and WASOsm to TSTo. Multicollinearity Prior to runn ing the multilevel models to predict TSTo within-person and between-person correlations were run between WASOsm and SOLsm variables to determine the extent to which they shared variances. Formal multicollinearity diagnostic procedures are not available for multilevel modeling. However, aim three of this study showed that WASOsm and SOLsm are collinear. Therefore, to contro l for their collinear relationshi p this diagnostic procedure will extract the shared variance from of SOLsm from WASOsm so that when they are individually added to the MLM the significant level will repres ent the unique contribution of each variable to the model. Within-Person Multicollinearity The within-person correlational an alysis re vealed significant collinearity between SOLsm and WASOsm (r = 0.-14, p < .05). Between-Person Multicollinearity The between-person correlationa l an alysis revealed significan t collinearity between SOLsm and WASOsm (r = 0.44, p < .01). Total Sleep Time (TST o) The intraclass correlation coefficient (ICC) for TSTo is 0.51. This indicates that, 49% of the overall variability in TSTo was within-person and 51% was between-person. For a complete listing of model parameters and estimates obtained at each step of the mo del building process see Table 5-7.
55 In the final MLM for TSTo, Day, = -0.74, t (111.74) = -1.99, p < .05, WASOsm, = -0.15, t (49.77) = -2.47, p = .02, and SOLsm, = 0.32, t (67.74) = 5.02, p < .001, were significant withinperson, level 1, predictors. At the between-person level, level 2, SOLsm, = 0.67, t (98.18) = 2.19, p = .03, was the only significant predictor of average TSTo. The model also contained a significant random effect of Day, = 2502.51, Walds Z = 6.67, p < .001, WASOsm, = 0.11, Walds Z = 2.13, p = .03, and SOLsm, = 0.10, Walds Z = 2.41, p = .02. This model explained approximately 11% of the within-person variance and 16% of the between-person variance. See Table 5-8 for a total listing of predictor estimates and significance levels.
56Table 5-1. Amount of withinand between-person variability Variable Sample standard deviation (SD) Individual standard devia tion (ISD) ISD / SD SOLsm 16.48 25.74 1.56 WASOsm 29.89 28.42 0.95
57Table 5-2. Amount of within and between-person variab ility by complaint status. Complaint Status Variable Sample St andard Deviation (SD) Individual St andard Deviation (ISD) ISD / SD No SOLsm 23.76 35.51 1.49 WASOsm 26.27 22.8 0.87 Yes SOLsm 12.17 21.90 1.80 WASOsm 34.29 43.28 1.26
58 Table 5-3. Steps taken in building the WASO s m multilevel model WASOsm Models AIC BIC -2LL -2LL df df s2 b s2 w r2 b r2 w (1) Null 4954.14 4969.80 4948.14 -101.42 -1.31 1.83 --(2) Time added 4938.83 4964.93 4928.83 19.31** 136.68 35.26 1.29 1.75 .02 .04 (3) SOLsm added 4811.11 4868.87 4811.11 117.7*** 102.48 -32.201.15 1.52 .11 -.13 Notes: AIC = Akaikes Information Criterion; BIC = Schwarzs Bayesian Criterion; -2LL = -2 log likelihood; -2LL = change in 2LL relative to preceding model; s2 b = unexplained intercept-related (b etween subjects ) variance; s2 w = unexplained residual-related (within subjects) variance; r2 b = between-subjects pseudo R-squared, an estimate of the amount of between subjects variance (estimated from null model) explai ned by fixed and random predictors; r2 w = within-subjects pseudo R-s quared, an estimate of the amount of within subjects vari ance (estimated from null model) explained by fixed and random predictors. *** Deviance is significant at the 0.001 level. ** Deviance is significant at the 0.01 level. Devi ance is significant at the 0.05 level.
59 Table 5-4. The relationship between SOLsm and WASOsm Fixed effects B SE df t p Predictor variable Within-person Day -0.03 0.01 129.50 -3.11 <.01 SOLsm -0.10 0.06 91.09 -1.78 .08 Between-person SOLsm 0.62 0.19 101.43 3.29 <.01 Random effects Covariance parameter estimate B SE Z p Within-person Day <0.01 <0.01 2.42 <.05 SOLsm 0.13 0.03 3.83 <.01 Within pseudo R2-.13 Between pseudo R2.11 Notes: Variables with a subscript sm indicate th ey were calculated by su btracting the subjective Seep Diary measure from the respective va riable measured by objective Actigraphy.
60Table 5-5. Steps taken in building the WASO s m multilevel model WASOsm Models AIC BIC -2LL -2LL df df s2 b s2 w r2 b r2 w (1) Null 14062.79 14084.86 14062.79 -103.00 -198.82 1007.27 --(2) Time added 14061.61 14087.93 14051. 6111.18 227.01 124.01173.36 981.60 .13 -.03 (3) SOLsm added 13342.36 13384.12 13326. 36725.25***172.55 54.46 182.45 839.37 .17 -.08 Notes: AIC = Akaikes Information Criterion; BIC = Schwarzs Bayesian Criterion; -2LL = -2 log likelihood; -2LL = change in 2LL relative to preceding model; s2 b = unexplained intercept-related (b etween subjects ) variance; s2 w = unexplained residual-related (within subjects) variance; r2 b = between-subjects pseudo R-squared, an estimate of the amount of between subjects variance (estimated from null model) explai ned by fixed and random predictors; r2 w = within-subjects pseudo R-s quared, an estimate of the amount of within subjects va riance (estimated from null model) explained by fixed and random predictors. *** Deviance is significant at the 0.001 level. ** Deviance is significant at the 0.01 level. Devi ance is significant at the 0.05 level.
61 Table 5-6. The relationship between SOLsm and WASOsm Fixed effects B SE df t p Predictor variable Within-person Day -0.74 0.21 204.24 -3.47 <.01 SOLsm -0.10 0.05 87.59 -2.01 <.05 Between-person SOLsm 0.21 0.05 98.09 3.98 <.01 Random effects Covariance parameter estimate B SE Z p Within-person Day 0.47 0.45 1.06 .29 SOLsm 0.13 0.03 4.02 <.01 Within pseudo R2-.08 Between pseudo R2.17 Notes: Variables with a subscript sm indicate they were calculated by subtra cting the subjective Seep Diary measure from the respective vari able measured by objective Actigraphy.
62Table 5-7. Steps taken in building the TSTo multilevel model TSTo Models AIC BIC -2LL -2LL df df s2 b s2 w r2 b r2 w (1) Null 15315.60 15331.3715309.60 -102.03 -2401.472522.99 --(2) Time added 15301.92 15328.2015291. 9217.68 124.40 22.372317.902316.36.08.03 (3) WASOsm added 14763.58 14805.3414747.58544.34***110.82 13.582203.652370.94.06.08 (4) SOLsm added 14680.95 14738.3614658.9588.63 110.16 0.66 2021.142252.68.11.16 Notes: AIC = Akaikes Information Criterion; BIC = Schwarzs Bayesian Criterion; -2LL = -2 log likelihood; -2LL = change in 2LL relative to preceding model; s2 b = unexplained intercept-related (b etween subjects ) variance; s2 w = unexplained residual-related (within subjects) variance; r2 b = between-subjects pseudo R-squared, an estimate of the amount of between subjects variance (estimated from null model) explai ned by fixed and random predictors; r2 w = within-subjects pseudo R-s quared, an estimate of the amount of within subjects vari ance (estimated from null model) explained by fi xed and random predictors. *** Deviance is signif icant at the 0.001 level. ** Deviance is sign ificant at the 0.01 level. Deviance is significant at the 0.05 level.
63 Table 5-8. Sleep misperception variables predicting TSTo Fixed effects B SE df t p Predictor variable Within-person Day -0.74 0.37 111.74 -1.99 <.05 WASOsm -0.15 0.06 49.77 -2.47 .02 SOLsm 0.32 0.06 67.74 5.02 <.01 Between-person WASOsm 0.09 0.18 98.58 0.51 .61 SOLsm 0.67 0.31 98.18 2.19 .03 Random effects Covariance parameter estimate B SE Z p Within-person Day 2502.51 374.96 6.67 .01 WASOsm 0.11 0.05 2.13 .03 SOLsm 0.10 0.04 2.41 .02 Within pseudo R2 .11 Between pseudo R2 .16 Notes: Variables with a subscript sm indicate they were calculated by subtra cting the subjective Seep Diary measure from the respective variable m easured by objective Actigraphy. The Subscript 'o' indicates the variable was m easured by objective Actigraphy.
64 CHAPTER 6 DISCUSSION Review of Findings Aim 1 and 2 Sleep m isperception is common but highly variable among older adults. The finding that older adults are almost as variable in WASOsm (ISD/SD = .95) and more variable in SOLsm (ISD/SD = 1.56) within-persons as between-persons is noteworthy, because it conflicts with the traditional view that sleep misperception is a consis tent trait like behavior isolated to patients with insomnia. Indeed, an ISD/SD value of 1.56 means that 85% of the total varian ce of the variable SOLsm across two weeks is found within-pers ons, and an ISD/SD value of .95 means that nearly 50% of the total variance in the variable WASOsm was within-persons. While, standard statistical techniques la rgely disregard within-person vari ability as error, herein it is argued that large amounts of within -person variability relative to between-person variability (i.e., ISD/SD > .50) may be mean ingful and predictable. Results from Aim 2 provide further evidence th at sleep misperception is not best explained by a between-persons phenomenon in older a dults. Some researchers argue that sleep misperception found among insomnia patients is limited to a subgroup making up about 25% of individuals with insomnia. From this perspective, that sleep misperception is limited to a subgroup of insomnia patients, we would expect to find larger amounts of between-persons variability than within-persons in a sample of individuals with sleep complaint. However, this is not what was found in this study. In fact, older ad ults with sleep compliant had between 65% and 75% more within-persons variability than between-persons variabili ty in the two sleep misperception variables tested in this study (i.e., SOLsm and WASOsm, respectively). These findings supplement previous findings from our lab in which we found that the number of days
65 overestimated SOL and WASO only occurred on average about 7 and 4 days respectively among older adult sleep complainers (Kay & McCrae, 2008) In addition, these results help explain the findings of a previous single night study which f ound that insomnia patients where just as likely to overas under-estimate SOL (V anable et al., 2000). The results of the present study suggest that the probability of older adults with insomnia overor under-estimating may be better predicted by within-person variability than by be tween-persons variability. Because on any given day individuals with sleep complaint are as likely to over-estimate SOL as under-estimate it, investigating the temporal or day-to-day factors that fluctu ate with sleep misperception may yield more valuable information about the cause of this phenomenon than investigating the more tonic factors such as personality or diagnostic categorization. Ultimately, studies that rely on a single night and averaging methods may not appr opriately represent th e construct of sleep misperception and certainly are not adequate to describe the longitudinal patterns of sleep misperception. The validity of sleep state misperception as a categorical diagnostic categorization may have limited validity among older adults. Aim 3 Though not previously validated, the convention of com bining average SOLsm and WASOsm seems to be supported by this study, because SOLsm and WASOsm co-occured on average. Since there was no interaction between complaint status and the between-persons relationship between SOLsm and WASOsm, we can conclude that rega rdless of whether there is a complaint or not, individual sleep mispercep tions are related on average. However, SOLsm and WASOsm do not necessarily co-occur on the same night. This finding, that the SOLsm and WASOsm shared a negative relationship on the daily level, adds to the appa rent discrepancy in findings between Mercer et al, (2002) who found that sleep misperceptions occurred during researcher probe throughout the night and Co ates et al. (1987) who found that sleep
66 misperception occurred only in the early part of the night and during spontaneous but not researcher induced awakenings later in the night. The major difference between these two studies is the location and the induction of arousal that may be associ ated with the laboratory as compared to the home environment. The former occurred in the laboratory while the later occurred in the home environment. Those indivi duals who where probed in the laboratory may have been more aroused throughout the night than those in the home environment. Because SOLsm and WASOsm did not co-occur daily, the sugg estion made by Perlis et al. (2001) that sleep misperception cannot be re duced to a general pess imistic bias imposed consistently on all aspects of sleep by individu als with sleep complaint is strengthened. Sleep misperception does not occur every day among sleep complainers, and SOLsm and WASOsm do not fluctuate in the same direction night-to-night In an additional analysis, sleep complaint was added as an additional predictor variable and the relationship between SOLsm and WASOsm was not found to be different by complaint stat us. Thus, sleep misperception may not be fundamentally different among sleep co mplainers and non-complainers beyond the frequency/severity. Aim 4 This analy sis revealed a complex relati onship between sleep misperception and TSTo. Specifically, SOLsm was significantly related to longer TSTo, while WASOsm was significantly related to shorter TSTo. One explanation for the di rectional difference in SOLsm and WASOsm on TSTo is that WASOsm is related to a more pervasive proble m of arousal leading to earlier wake times and/or reduced sleep during the night while SOLsm may be related to a more phasic arousal problem occurring primarily during the sleep onset period. Vanable et al (2000) argued that poorer sleep quality may motivate in dividuals to exaggerate their sleep problems, thus explaining sleep misperception. Contrary to their theory, this analysis clearly suggests that when
67 individuals TSTo is greater than usual, they ar e more likely to overestimate SOLsm and when their sleep is shorter than usual, they are more likely to overestimate WASOsm, regardless of complaint status. We argue that SOLsm may represent a form of loca lized sleep deprivation that reduces the likelihood of WASOsm and extends the length of sleep, similar to the effects of global sleep deprivation. Summary of Results These analyses suggest that though sleep m isp erception is common among older adults, it is not limited to individuals with sleep complain t. Identification of sleep perceiver subtypes of insomnia may be the result of limited observa tions or averaging techniques that overlook individuals highly variable sleep behavior. The common view that sleep misperception represents an exaggeration bias does not have su pport. This study suggests that older adults with sleep misperception follow patterns more simila r to sleep deprivation than exaggeration. The validity of sleep misperception as a distinct sleep disorder that is quant itatively different in patterns among insomnia patients and normal sleeper s should be reviewed. Additional studies are needed to quantify the in traindividual variability of sleep misperceptions in diagnosed insomnia and sleep state misperception patients. Study Limitations This secondary analysis study was based on a convenience sam ple a nd not a scientific experimental study. Limitations of the present stu dy include restricted gene ralizability of results. The results of this study may not generalize be yond community dwelling older adults. There are other methodological concerns as well. The use of a convenience sample restricted the diversity of the sample. The participants were primarily European Caucasian, a nd college educated. The homogeneity of the sample prevents reliable generalization to younger or diverse populations. Additionally, individuals were ex cluded from the study if they presented with sleep disorders
68 other than insomnia (e.g. sleep apnea, periodic leg movements). Approximately one-half of the elderly population experience one or both of thes e conditions (Ancoli-Is rael, Kripke, Mason, & Kaplan, 1985). A New Model of Sl eep Misperception One prom inent theory of sleep argues that ther e are two major sleep systems, or processes (Borbely, 1982). The circadian or C-process is an endogenous process imposing sleep on the brain based on species specific bi ological rhythms while the S-pr ocess or homeostatic process progressively imposes sleep on the brain proporti onal to the amount of previous wakefulness. These systems function on both global and local levels of brain functioning. However, the body/mind and sleep/wake false dichotomies compli cate the discussion of sleep misperception. First, the idea that the mind is orthogonal from wh at occurs in the brain is scientifically invalid. All mental events correspond to brain events. Ho wever, to reduce the mind to the sum of the parts of the brain is equally, though philosophically, invalid. From the perspective of the brain, for example, there is no meaning in the perception of sleep as wakefulness. This false dichotomy is apparent in delineations of cognitive ar ousal as a function of mental arousal while physiological arousal is supposed to be a func tion of body arousal. However, cognitive arousal corresponds to activation of neurons in executive centers of the brain while physiological arousal corresponds to activation of the descending sympathetic nervous sy stem in the brain. Both are physiological and both occur in distant, though in terconnected, parts of th e brain (discussed in detail below). All that matters in relationship to sleep misperception is whether the source of arousal activates consciousness centers of the brain while global sleep has set in. Second, sleep and wake are not dichotom ous global state of brain, physiology, or consciousness. The sleeping brain is dynamic and ne ver in a ubiquitous pattern of neuroactivity in all areas of the brain at any time (Rot h, Achermann, & Borbely, 1999; Werth, Achermann, &
69 Borbely, 1997). Though the use of EEG has become the gold standard for determining sleep onset, early sleep researchers warned against the sole use of EEG as a measure of sleep onset (Kleitman, 1963). Physiologically, transitions between sleep and wake occur on a continuum in space and time with no clea r either/or. While arbitrary cutoffs have been entrenched in sleep research, these delineations are arbitrar y by all accounts (Tryon, 2004). In healthy young populations, self-report measures of sleep onset tend to occur later than EEG defined sleep and EEG defined sleep tends to o ccur later than behavioral in dices of sleep (Tryon, 2004) and currently there is no evidence that this pattern holds in diverse patient populations. Based on the localized model of sleep, it is en tirely possible for this typical pattern to occur in any order. Moreover, even though EEG definitions of sleep were created based on be havioral correlates, EEG definitions of sleep and beha vioral indices often disagree. One study found that dogs given atropine resulted in EEG defined sleep while the dog was clearly awake and active (Wikler, 1952). Certain brain damage in humans has been shown to induce EEG defined sleep while the individuals are observably awak e (Hamoen, 1954 cited in (Hauri & Olmstead, 1983). Actigraphy may be a valid indication that the afferent moto r centers of the brain have gone to sleep, while EEG may indicate that the neocortex has fallen asl eep, and self-report may indicate the point in time that the consciences centers of the brain ha ve fallen asleep. People w ith sleep misperception may have a longer transition period with a unique brain dynamic unlike hea lthy sleep onset such that behavioral and EEG based in dices reflect a sleep state while the persons cognitive centers of the brain remain awake and the pers on remains aware of self and time. The idea that consciousness switches off at slee p onset is as untenable as the idea, almost ubiquitously held (Saper, Chou, & Scammell, 2001), that there is a sleep switch inducing the dichotomy between sleep and wake. Nonetheless, assumptions, highlighted in the sleep switch
70 theory remain highly accepted and popular among sl eep experts. Levels of conscious awareness, attention to and processing of the environment, similar to process observed in wakefulness, are on a continuum which corresponds to the level and depth of sleep of areas of the brain involved (Czisch et al., 2002). For example, during sleep some level of attention is attuned to external stimuli and depending on the contex t of stimuli sleep will continue or terminate. Specifically, the participants name induced awakenings more successfully than a neutral stimulus (Perrin, Garcia-Larrea, Mauguiere, & Bastuji, 1999). Ex plicit to sleep misper ception, between 27 and 50% of healthy sleepers reported not having slep t minutes after sleep spindles emerged and between 10 and 5% of these individuals report ed wakefulness during stage 2 and stages 3-4 sleep, respectively (reviewed in Bonnet & Moor e, 1982). The hypothalamic sleep switch theory must be reconsidered and may more appropriately be called the hypothalamic sleep promoting theory that interacts with other sleep promoti ng and arousal systems in the brain. Even early sleep researchers acknowledged that cogi tation occurs during sleep (Mullin, 1938). Unfortunately, contemporary sleep experts s eem unable to embrace the idea that the incongruence between a participan ts self-reported experience and EEG defined states is the product of limited recording reso lution. Evidence suggests that EEG is unable to measure the physiological state of the brain areas involved in consciousne ss. As Sewitch (1984) argued, When a sleeper is awakened at some point durin g the night and asked to evaluate his or her subjective state prior to the awakening, he or she is confronted w ith a decision between two alternatives, sleep, or wakefulness. Conse quently, it seems appropriate to consider the issue of perceiving sleep and wakefulness in the context of a decision problem. There are three elements essential to any decision problem [One of these essential elements is that] there must be two states of the world and in the case of deciding about ones subjective state during the sleep period, those two states are wakefulness versus sleep. (p. 244-245) In fact, sleep and wake are not dichotomous; th erefore, the decision need not be wake vs. sleep. Although teaching patients how to judge what the EEG defines as sleep is effective in treating sleep misperception (Dow ney & Bonnet, 1992), teaching pati ents to accurately report
71 their experience in terms of EEG cutoffs does no t necessarily inform why they needed this training in the first place, nor does it necessar ily mean that their experience has substantively changed. Ignoring the patient se lf-reported state may prevent th e best treatments for sleep misperception from being formed. This techniqu e may indirectly improve insomnia by teaching patients that most of the brai n is obtaining restful sleep, thus reducing anxiety. However, it probably does not directly mean that the whole br ain, including the perceptual areas of the brain, is receiving healthy sleep. Implications for Sleep Research, Diagnoses, and Treatments It is argued that sleep m isper ceptions may result in two ways. First, simultaneously high activation in the sleep promoting and arousal systems of the brain would allow self-awareness to occur during lighter stages of sleep. Second, low localized sleep drive in consciousness centers of the brain in the presence of high global sleep drive may allow is olated areas to remain active during EEG defined sleep. These two possible path ways to sleep misperception while having a similar symptoms (perception of EEG defined sleep as wakefulness), may be different in their consequences. The former would induce maintained activation in only isolat ed areas of the brain, representing a form of localized sleep deprivation in only those a ffected areas of the brain, while the later would be related to a more pervasive flattening of the S-pr ocess in localized areas of the brain. The treatment of sleep misperception re sulting from hyperarousal may employ relaxation techniques. Conversely, to treat sleep misperception rela ted to a localized S-process deficiency may involve enhancing daytime activation of affect ed brain areas. This would force sleep dept to accumulate in those areas, thus, enhancing slee p in those areas of the brain and consolidating their sleep processes with the whole brains sl eep/wake pattern. In kind, the mechanisms of insomnia onset may come through either of these forms, suggesting that at least two subtypes of insomnia seem logical: the first resulting fr om hyperarousal, the ot her from a S-process
72 deficiency. It is time that re searchers begin to focus more on the mechanisms of insomnia to determine its subtypes, rather than exclusiv ely studying diagnostic categ ories developed from clinically observed symptom constellations. Sleep misperception has been at the center of the sleep disorder debate. The validity of all insomnia studied and the diagnosis of insomnia its elf was at one time considered to be threatened by the discovery of sleep misper ception (Borkovec et al., 1979). Indeed, one may wonder, What if there are no global sleep problems among insomnia patients? or What will become of all the studies that include a large subgr oup of participants, as many as half all insomnia patients, who do not have global sleep deficien cies? Since that time, sleep state misperception has been accepted by most researchers and clinicians. More over, its systematic validation has become a fixed pursuit of the field. Regrettably, these fears seem to have led the field to embrace a system of diagnostic categories that re quire retrospective va lidation at the expense of scientific exploration and prospective progr ess in understanding insomnia and its relationship to sleep misperception. One explanation offered by Borkovec ( 1981) that should be reconsidered is that insomnia patients may.base their evaluation of sleep on a different set of internal and/or external cues relative to good sl eepers (p.609). A slight modificat ion to this statement may be more accurate: that [hyperaroused individuals with or without a sl eep complaint] may base their evaluation of sleep on a different se t of internal and/or external cues relative to [individuals that are not hyperaroused]. As sleep experts, and societ y as a whole, continue to define sleep as an either/or state, individuals who experience sleep misperception may have difficulty determining which category their subjective experience belo ngs (Lamarche & Ogilvie, 1997) since it may not fit in either. The validity of sleep state misperception as a distinct disorder has little support. However, identification of sleep misperception in a patient shoul d receive treatment even if
73 objective measures are absent, because sleep mi sperception may be effectively treated. Unlike Borkovek (1981) who once argued that the presence of localized arousal events may require a revision of the EEG criteria for sleep, at least for some clinical populations (p608), we conclude that altering the criteria of a global measure based on a localized event may not adequately capture the localized event. That is, this a pproach may either overor under-approximate the impact of localized activity on a global scale. The results of this study shows that there is much knowledge that can be gained by determining th e amount of discrepancy between the perception of sleep onset and the EEG gl obal assessment on sleep onset on both the daily level and the within-persons level. Treatment implications Sleep m edicine has made dramatic headway in the diagnosis and treatment of sleep problems. Yet, the development of a univers ally accepted definition of insomnia remains illusive. Moreover, many insomnia treatments employ a shotgun approach, that is, giving multiple interventions and hoping one works. Often th is approach is effective; however, in some cases it is not as efficient or effective as desired. Sleep medicine is at the point where understanding the pathophysiology of insomnia, and sleep in disord ers highly related to insomnia more generally, is required so that future tr eatments can be developed (Dang-Vu et al., 2007; Mahowald & Schenck, 2005). Most importantly, sleep research and medicine must accept that sleep is not a whole brain event (Krueger & Ob al, 2003). In addition, re gional abnormalities in sleep architecture need to be better underst ood through research, and spec ifically targeted in treatment. More importantly, localized sleep ab normalities, in the form of extended activation during sleep, can be couched in the framework of localized sleep depriv ation. Like global sleep deprivation, localized sleep depr ivation may predispose individuals to a host of mental and physical health problems. Given th e hypothesis that isolated areas of the brain remain active, or
74 are prevented from sleeping deep ly with the rest of the brai n during sleep mi sperception, these areas may be subjected to a form of localized sleep deprivation. Like global sleep deprivation (Durmer & Dinges, 2005), localized sleep depriv ation may predispose individuals to problems related to the areas of the brain that are affected in sleep misperception. Moreover, sleep misperception may explain the high comorbidity ra tes of hyperarousal rela ted health conditions and insomnia. For example, between 50-70% of individuals with chronic pain commonly report sleep problems (Pilowsky, Crettenden, & Townley, 1985). The high co-morbidity between these health problems hints at a common pathophysiology, but the relati onship between these conditions remains undetermined. There may be three potential fact ors related to sleep misperce ption. If the combination of these factors is better understood new treatments may be developed to improve sleep locally and, thus, prevent or improve potential problems. Fo r example, if there is a localized S-process deficiency in only the prefrontal cortex, a treat ment such as biofeedback may be developed to enhance the sleep need of that area. In another situation, if there is in creased arousal in the anterior cingulated, new treatments may be ab le to reduce localized arousal behaviorally, allowing the individual to sleep more deeply at night with the rest of the brain. Thus, both potentially eliminating sleep mispercep tion and localized sleep deprivation. Future Directions This study highlights the needed and utility of IIV and MLM analyses in sleep research. Sleep researchers need to look beyond single ni ght and averaging research methods in sleep research broadly and sleep misperception resear ch specifically. Sleep misperception, as with insomnia, appears to be periodic rather than c onsistent among sleep compla iners, not only from night-to-night but also throughout the night. Moreover, it is interesting that over two weeks average SOLsm correlated with average WASOsm, yet, this was not true on a nightly basis.
75 Additional studies are needed to replicate these findings in yo unger populations and in diagnostic groups. More importantly, neuroimaging studies that combine functional/structural imaging, PSG measures, and self-report in concert ar e needed to uncover the biopsychophysiological mechanism of sleep misperceptions. Regional sleep theories argue that sleep is not a whole brain event, but that each region of the brain has its own sleep patterns and needs (K attler, Dijk, & Borbely, 1994; Krueger & Obal, 1993, 2003; Krueger, Obal, & Fang, 1999; Lavie, 1993; Mahowald & Schenck, 1992; Mukhametov, Supin, & Poliakova, 1984; Vyazovs kiy, Borbely, & Tobler, 2000; Vyazovskiy, Borbely, & Tobler, 2002; Yasuda, Yasuda, Brown, & Krueger, 2005; Yoshida et al., 2004). Sleep and wakefulness are not dichotomous globa l states. Although there is ample evidence of this, few have embraced this idea and integrated it into how they think about and do sleep research. Indeed, this idea has b een limited to basic animal resear ch and theory (Krueger & Obal, 2003; Mahowald & Schenck, 1992), few healthy hu man studies (Czisch et al., 2004; Kattler et al., 1994; Kaufmann et al., 2006; Maquet, 2000; Wehr le et al., 2005), and recently, a handful of neuroimaging studies of selected sleep disorders. Several mari ne mammals and bird species provide evidence that sleep and its functions can occur in isolated areas of the brain simultaneously while other areas continue to pe rform wake-like brainwaves and functions, such as attending to the environm ent (Mukhametov, Supin, & Poliakova, 1984). Even in humans, isolated events of activity i ndicative of wakefulness occur du ring sleep (Dement & Kleitman, 1957) and isolated events of slow wave activity indicative of sleep oc cur during wakefulness (Durmer & Dinges, 2005). Both of these scenario s are far more common than realized. When one considers that the brain is a complex organ and is never in a ubiquito us state of identical activity in all areas of the brain during wakefulness, it seems odd that sleep is treated as a whole
76 brain event. For example, most EEG sleep resear ch relies on as few as one EEG electrode that provides summations of brain wave activity, partic ularly activity occurring on the surface of the cortex. While few would dispute that daytime br ain activity is localized during wakefulness, even fewer have considered that such localiza tion can occur during sleep. Indeed, EEG research with multiple leads (Roth et al., 1999) a nd neuroimaging studies support this hypothesis (Kaufmann et al., 2006). It is cl ear that there is a continuum of cumulative global states between sleep/wake stages (Kaufmann et al., 2006). Furtherm ore, given that sleep is not a whole brain event, there is little support for the idea that sleep can only oc cur in the complete absence of consciousness (Mercer et al., 2002). Consciousness during sleep would result if areas of the brain responsible for consciousness remain active while other areas deact ivate into slow wave activity. The prefrontal cortex, and the anterior cingulate cortex (Bush, Luu, & Posner, 2000) have been posited as the seat of consciousness. Given that sleep misperception is the subjective perception of wakefulness during sleep, undet ectable to most EEG procedures (i.e., sleep misperception), may represent extended microarousal in th ese consciousness cente rs during nonrapid eye movement sleep (NREMs). Sleep misperception may represent a form of localized sleep deprivation resulting in the dysre gulation and dysfunction in these isolated and related areas of the brain. This may be the source of sleep diso rders and mental and phys ical health problems. Specifically, we posit that sleep misperception is the result of concurrent activation of both the sleep system (forcing EEG defined sleep to occu r) and the ascending arousal system preventing selected areas from sleeping with the rest of the brain. While a microstructure indication of localized activation in sleep misperception has been identified employi ng EEG (Mercer et al., 2002) the location of this sleep problem has not been studied to our knowledge. Cumulative global states are simply the amalgamation of di verse brain functions a nd physiological activity.
77 The goal of research in the future should be aimed at defining the mechanisms and factors involved that predict the days when sleep misp erception will occur or not. These studies may help determine whether arousal directly or indir ectly is related to slee p misperception and how a local of global S-process deficiency may be involved. This line of research may finally be able to confirm or refute the uncertainty of Borkovec (1981) when he said, The experience of wakefulness during sleep is indeed a remarkable phenomenon. Although the effects may be ultimately determined to be due to artifact or gross EEG definitions of sleep, its elucidation should contribute to an understandin g of some of the contributors to the experience of insomnia (p 609). The findings from the present study co ntribute to decades of research aimed at understanding sleep misperceptions role in insomnia and may be the beginning steps in a new avenue of insight to this problem.
78 APPENDIX A SLEEP DIARY SLEEP DIARY ID# ______________ Week 1 ___ Week 2___ Please answer the following questionnaire WHEN YOU AWAKE IN THE MORNING Enter yesterday's day and date and provide the information to describe your sleep the night before. Definitions explaining each line of the questionnaire are given below. EXAMPLE yesterday's day yesterday's date TUES 10/14/97 day 1 day 2 day 3 day 4 day 5 day 6 day 7 1. NAP (yesterday) 70 2. BEDTIME (last night) 10:55 3. TIME TO FALL ASLEEP 65 4. # AWAKENINGS 4 5. WAKE TIME (middle of night) 110 6. FINAL WAKE-UP 6:05 7. OUT OF BED 7:10 8. QUALITY RATING* 2 9. BEDTIME MEDICATION (include amount & time) Halcion 0.25 mg 10:40 pm *Pick one number below to indicate your overall QUALITY RATING or satisfaction with your sleep. 1. ver y p oor 2. p oor 3. fair 4. g ood 5. excellent
79 APPENDIX B HEALTH SURVEY HEALT H SURVEY Please PRINT and Supply A A L L L L Information ID# Height _________ Weight __________ Race __________________ 1. Do you have a sleep problem? yes or no If yes, describe (e.g., troubl e falling asleep, long or frequent awakenings, sleep apnea): ___________________________________________________________________________ If yes, on average, how many nights pe r week do you have this problem? ________________ How long have you had this sl eep problem? ________years ________months 2. Please indicate whether you or your bed partner have not iced any of the following: Are you a heavy snorer? yes no Do you have difficulty breathing or gasp for breath during sleep? yes no Do your legs jerk frequently during sleep or do th ey feel restless before sleep onset? yes no Do you have sleep attacks during the day or paralysis at sleep onset? yes no If yes to any of the questions under #2, please explain and indicate how often symptoms occur: ________________________________________________________________________ ________________________________________________________________________ 3. Indicate with a check mark if you have the following health problems, and put the number of years youve had each problem: Yes Years ___ _______ Heart disease ___ ________ Cancer ___ ________ AIDS ___ ________ High blood pressure ___ ________ Neurological disease (ex: seizures, Parkinsons) ___ ________ Breathing Problems (ex: asthma, emphysema) ___ ________ Urinary problems (ex: ki dney disease, prostate problems) ___ ________ Diabetes ___ ________ Chronic Pain (ex: ar thritis, back pain, migraines) ___ ________ Gastrointestinal (ex: stomach, irritable bowels, ulcers) 4. Please list any mental health disorders you ha ve and the number of years youve had the disorder(s)
80 (OVER) 5. List any other health problems you have (and the number of years youve had the problem). _____________________________________________________________________________ 6. Medical and mental health di sorders may disrupt sleep. Medi cation may also disturb sleep. Please list any disorder or medica tion that affects your sleep and describe how it affects sleep. ______________________________________________________________________________ ______________________________________________________________________________ 7. List ALL medications taken within the past month, the fr equency with which they are taken (e.g., daily, 3 times a day, weekly), time of day, and the purpose of the medication. Medicine Frequency Time of Day Purpose a. __________________________________________________________________________ b. __________________________________________________________________________ c. __________________________________________________________________________ d. __________________________________________________________________________ e. __________________________________________________________________________ f. __________________________________________________________________________ g. __________________________________________________________________________ 8. List ALL vitamins taken within the past month, the frequency with which they are taken (e.g., daily, 3 times a day, weekly), time of day, and the purpose of the medication Vitamin Frequency Time of Day Purpose a. __________________________________________________________________________ b. __________________________________________________________________________ c. __________________________________________________________________________ d. __________________________________________________________________________
81 9. On average, how many alcoholic drinks do you drink per week? ____________ 10. On average, how many cigare ttes do you smoke per day? _______________ 11. On average, how many caffeinated drinks do you have per day? __________ 12. What is your highest level of education? 13. If you have a spouse, what is his or her highest level of education?
82 REFERENCE LIST Affleck, G., Tennen, H., Urrows, S., Higgins, P., Abeles, M., Hall, C., et al. (1998). Fibrom yalgia and women's pursuit of pers onal goals: a daily process analysis. Health Psychol, 17(1), 40-47. Affleck, G., Urrows, S., Tennen, H., Higgins, P., & Abeles, M. (1996). Sequential daily relations of sleep, pain intensity, and attention to pain among women with fibromyalgia. Pain, 68(2-3), 363-368. Ancoli-Israel, S., Cole, R., Alessi, C., Chambers M., Moorcroft, W., & Pollak, C. P. (2003). The role of actigraphy in the study of sleep and circadian rhythms. Sleep, 26(3), 342-392. Attarian, H. (2007). Disturbed perception of wakefulness within sl eep: a new sleep disorder or the extreme end of a spectrum. Sleep Med, 8 (2), 103-104. Attarian, H. P., Duntley, S., & Brown, K. M. (2004). Reverse sleep state misperception. Sleep Med, 5(3), 269-272. Benoit, O., & Aguirre, A. (1996). Homeostatic a nd circadian aspects of sleep regulation in young poor sleepers. Neurophysiol Clin, 26 (1), 40-50. Bliwise, D. L., Friedman, L., & Yesavage, J. A. (1993). Depression as a confounding variable in the estimation of habitual sleep time. J Clin Psychol, 49 (4), 471-477. Bonnet, M. H., & Arand, D. L. (1992). Caffeine us e as a model of acute and chronic insomnia. Sleep, 15(6), 526-536. Bonnet, M. H., & Arand, D. L. (1995). 24-Hour metabolic rate in insomniacs and matched normal sleepers. Sleep, 18 (7), 581-588. Bonnet, M. H., & Arand, D. L. ( 1997a). Hyperarousal and insomnia. Sleep Med Rev, 1 (2), 97108. Bonnet, M. H., & Arand, D. L. (1997b). Physiologi cal activation in patients with Sleep State Misperception. Psychosom Med, 59 (5), 533-540. Bonnet, M. H., & Moore, S. E. (1982). The threshold of sleep: perception of sleep as a function of time asleep and auditory threshold. Sleep, 5 (3), 267-276. Borbely, A. A. (1982). A two process model of sleep regulation. Hum Neurobiol, 1 (3), 195-204. Borbely, A. A. (1987). The S-deficiency hypothesis of depression and the two-process model of sleep regulation. Pharmacopsychiatry, 20(1), 23-29. Borkovec, T. D. (1982). Insomnia. J Consult Clin Psychol, 50 (6), 880-895.
83 Borkovec, T. D., Grayson, J. B., O'Brien, G. T., & Weerts, T. C. (1979). Relaxation treatment of pseudoinsomnia and idiopathic insomnia: an electroencephalographic evaluation. J Appl Behav Anal, 12 (1), 37-54. Borkovec, T. D., Lane, T. W., & VanOot, P. H. (1981). Phenomenology of sleep among insomniacs and good sleepers: wakefulness experience when cortically asleep. J Abnorm Psychol, 90(6), 607-609. Bush, G., Luu, P., & Posner, M. I. (2000). Cognitive and emotional influences in anterior cingulate cortex. Trends Cogn Sci, 4 (6), 215-222. Carskadon, M. A., Dement, W. C., Mitler, M. M., Guilleminault, C., Zarcone, V. P., & Spiegel, R. (1976). Self-reports versus sleep laborat ory findings in 122 drug-free subjects with complaints of chronic insomnia. Am J Psychiatry, 133 (12), 1382-1388. Chambers, M. J., & Kim, J. Y. (1993). The role of state-trait anxiety in insomnia and daytime restedness. Behav Med, 19(1), 42-46. Coates, T. J., George, J. M., Killen, J. D., Marchini, E., Hamilton, S., & Thorensen, C. E. (1981). First night effects in good sleepers and sleep-maintenance insomniacs when recorded at home. Sleep, 4(3), 293-298. Cole, R. J., Kripke, D. F., Gruen, W., Mullane y, D. J., & Gillin, J. C. (1992). Automatic sleep/wake identification from wrist activity. Sleep, 15 (5), 461-469. Coleman, R. M., Roffwarg, H. P., Kennedy, S. J., Guilleminault, C., Cinque, J., Cohn, M. A., et al. (1982). Sleep-wake disorders based on a polysomnographic diagnosis. A national cooperative study. Jama, 247(7), 997-1003. Craske, M. G., & Craig, K. D. (1984). Musical performance anxiety: the three-systems model and self-efficacy theory. Behav Res Ther, 22(3), 267-280. Czisch, M., Wehrle, R., Kaufmann, C., Wetter, T. C., Holsboer, F ., Pollmacher, T., et al. (2004). Functional MRI during sleep: BOLD signal decreases and their electrophysiological correlates. Eur J Neurosci, 20 (2), 566-574. Czisch, M., Wetter, T. C., Kaufmann, C., Pollmac her, T., Holsboer, F., & Auer, D. P. (2002). Altered processing of acoustic stimuli during sleep: redu ced auditory activation and visual deactivation detected by a combined fMRI/EEG study. Neuroimage, 16(1), 251258. Dang-Vu, T. T., Desseilles, M., Petit, D., Mazza, S., Montplaisir, J., & Maquet, P. (2007). Neuroimaging in sleep medicine. Sleep Med, 8 (4), 349-372. Davidson, L. M., Fleming, R., & Baum, A. (1987). Chronic stress, catecholamines, and sleep disturbance at Three Mile Island. J Human Stress, 13(2), 75-83.
84 de Souza, L., Benedito-Silva, A. A., Pires, M. L., Poyares, D., Tufik, S., & Calil, H. M. (2003). Further validation of actigraphy for sleep studies. Sleep, 26(1), 81-85. Dement, W., & Kleitman, N. (1957). Cyclic varia tions in EEG during slee p and their relation to eye movements, body motility, and dreaming. Electroencephalogr Clin Neurophysiol, 9(4), 673-690. Diagnostic and Statistical Manual of Mental Disord ers, Fourth Edition, Text Revision (DSM-IVTR) (2000). Washington, DC: Ameri can Psychiatric Association. Dorsey, C. M., & Bootzin, R. R. (1997). Subjective and psychophysiologic insomnia: an examination of sleep tendency and personality. Biol Psychiatry, 41 (2), 209-216. Downey, R., 3rd, & Bonnet, M. H. (1992). Training subjective insomniacs to accurately perceive sleep onset. Sleep, 15(1), 58-63. Drewes, A. M., Gade, K., Nielsen, K. D., Bjerrega rd, K., Taagholt, S. J., & Svendsen, L. (1995). Clustering of sleep electroence phalographic patterns in patients with the fibromyalgia syndrome. Br J Rheumatol, 34 (12), 1151-1156. Drewes, A. M., Nielsen, K. D., Taagholt, S. J ., Bjerregard, K., Svendsen, L., & Gade, J. (1995). Sleep intensity in fibromyalgia: focus on the microstructure of the sleep process. Br J Rheumatol, 34(7), 629-635. Drummond, S. P., Smith, M. T., Orff, H. J., Chengazi, V., & Perlis, M. L. (2004). Functional imaging of the sleeping brain: review of findings and implications for the study of insomnia. Sleep Med Rev, 8 (3), 227-242. Durmer, J. S., & Dinges, D. F. (2005). Neurocognitive consequences of sleep deprivation. Semin Neurol, 25(1), 117-129. Edinger, J. D., & Fins, A. I. (1995). The dist ribution and clinical si gnificance of sleep time misperceptions among insomniacs. Sleep, 18(4), 232-239. Edinger, J. D., Fins, A. I., Glenn, D. M., Sulliva n, R. J., Jr., Bastian, L. A., Marsh, G. R., et al. (2000). Insomnia and the eye of the beholder: are there clinical mark ers of objective sleep disturbances among adults with a nd without insomn ia complaints? J Consult Clin Psychol, 68(4), 586-593. Edinger, J. D., Fins, A. I., Goeke, J. M., McMill an, D. K., Gersh, T. L., Krystal, A. D., et al. (1996). The empirical identifica tion of insomnia subtypes: a cluster analytic approach. Sleep, 19(5), 398-411. Edinger, J. D., Fins, A. I., Sullivan, R. J., Jr., Marsh, G. R., Dailey, D. S., Hope, T. V., et al. (1997a). Do our methods lead to insomniacs' madness?: Daytime testing after laboratory and home-based polysomnographic studies. Sleep, 20 (12), 1127-1134.
85 Edinger, J. D., Fins, A. I., Sullivan, R. J., Jr., Marsh, G. R., Dailey, D. S., Hope, T. V., et al. (1997b). Sleep in the laboratory and sleep at home: comparis ons of older insomniacs and normal sleepers. Sleep, 20 (12), 1119-1126. Edinger, J. D., & Krystal, A. D. (2003). Subtyping primary insomnia: is sleep state misperception a distinct clinical entity? Sleep Med Rev, 7 (3), 203-214. Edinger, J. D., Marsh, G. R., McCall, W. V., Erwin, C. W., & Lininger, A. W. (1991). Sleep variability across consecutive nights of home monitoring in older mixed DIMS patients. Sleep, 14(1), 13-17. Edinger, J. D., Wohlgemuth, W. K., Radtke, R. A., Marsh, G. R ., & Quillian, R. E. (2001). Does cognitive-behavioral insomnia therapy alter dysfunctional be liefs about sleep? Sleep, 24(5), 591-599. Eriksen, H. R., & Ursin, H. (2004). Subjective h ealth complaints, sensitization, and sustained cognitive activation (stress). J Psychosom Res, 56 (4), 445-448. Espie, C. A. (2002). Insomnia: conceptual issues in the development, persistence, and treatment of sleep disorder in adults. Annu Rev Psychol, 53 215-243. Espie, C. A., Inglis, S. J., Tessier, S., & Ha rvey, L. (2001). The clinical effectiveness of cognitive behaviour therapy for chronic insomnia: implementation and evaluation of a sleep clinic in general medical practice. Behav Res Ther, 39 (1), 45-60. Fichten, C. S., Creti, L., Amsel, R., Bailes, S., & Libman, E. (2005). Time estimation in good and poor sleepers. J Behav Med, 28 (6), 537-553. Ford, D. E., & Kamerow, D. B. (1989). Epid emiologic study of sleep disturbances and psychiatric disorders. An opportunity for prevention? Jama, 262(11), 1479-1484. Frankel, B. L., Coursey, R. D., Buchbinder, R., & Snyder, F. (1976). Recorded and reported sleep in chronic primary insomnia. Arch Gen Psychiatry, 33 (5), 615-623. Freedman, R. R. (1986). EEG power sp ectra in sleep-onset insomnia. Electroencephalogr Clin Neurophysiol, 63 (5), 408-413. Freedman, R. R., & Sattler, H. L. (1982). Physiol ogical and psychological factors in sleep-onset insomnia. J Abnorm Psychol, 91 (5), 380-389. Harvey, A. G. (2000). Pre-sleep cognitive activity: a comparison of sleep-onset insomniacs and good sleepers. Br J Clin Psychol, 39 ( Pt 3) 275-286. Harvey, A. G. (2002a). A cognitive model of insomnia. Behav Res Ther, 40(8), 869-893.
86 Harvey, A. G. (2002b). Identifying safety behaviors in insomnia. J Nerv Ment Dis, 190 (1), 1621. Harvey, A. G., & Greenall, E. (2003). Catastrophic worry in primary insomnia. J Behav Ther Exp Psychiatry, 34 (1), 11-23. Harvey, A. G., Tang, N. K., & Browning, L. (2 005). Cognitive approaches to insomnia. Clin Psychol Rev, 25 (5), 593-611. Hauri, P. (1969). The influence of ev ening activity on th e onset of sleep. Psychophysiology, 5 (4), 426-430. Hauri, P., & Olmstead, E. (1983). What is the moment of sleep onset for insomniacs? Sleep, 6(1), 10-15. Hauri, P. J., & Wisbey, J. (1992). Wrist actigraphy in insomnia. Sleep, 15 (4), 293-301. Haynes, S. N., Adams, A., & Franzen, M. (1981) The effects of presle ep stress on sleep-onset insomnia. J Abnorm Psychol, 90 (6), 601-606. Hoch, C. C., Reynolds, C. F., 3rd, Kupfer, D. J ., Berman, S. R., Houck, P. R., & Stack, J. A. (1987). Empirical note: self-re port versus recorded sl eep in healthy seniors. Psychophysiology, 24 (3), 293-299. Horne, J. A., Pankhurst, F. L., Reyner, L. A ., Hume, K., & Diamond, I. D. (1994). A field study of sleep disturbance: effects of aircra ft noise and other factors on 5,742 nights of actimetrically monitored sleep in a large subject sample. Sleep, 17(2), 146-159. Houts, A. C. (2000). Fifty years of psychiat ric nomenclature: reflections on the 1943 War Department Technical Bulletin, Medical 203. J Clin Psychol, 56 (7), 935-967. Houts, A. C. (2001). The diagnos tic and statistical manual's new white coat and circularity of plausible dysfunctions: res ponse to Wakefield, part 1. Behav Res Ther, 39 (3), 315-345. Johns, M. W., Gay, T. J., Masterton, J. P., & Br uce, D. W. (1971). Relationship between sleep habits, adrenocortical activity and personality. Psychosom Med, 33 (6), 499-508. Karacan, I., Thornby, J. I., Anch, M., Booth, G. H ., Williams, R. L., & Salis, P. J. (1976). Doserelated sleep disturbances induced by coffee and caffeine. Clin Pharmacol Ther, 20 (6), 682-689. Kattler, H., Dijk, D. J., & Borbely, A. A. (1994). Effect of unilateral somatosensory stimulation prior to sleep on the sleep EEG in humans. J Sleep Res, 3 (3), 159-164.
87 Kaufmann, C., Wehrle, R., Wetter, T. C., Holsboer, F., Auer, D. P., Pollmacher, T., et al. (2006). Brain activation and hypothalamic functiona l connectivity during human non-rapid eye movement sleep: an EEG/fMRI study. Brain, 129(Pt 3), 655-667. Kay, D. B., McCrae, C. S., & Rowe, M (2008). Sl eep Misperception and Sleep Complaints in an Older Adult Population Sleep 28, A256. Kiley, J. P. (1999). Insomnia research and future opportunities. Sleep, 22 Suppl 2, S344-345. Krueger, J. M., & Obal, F. (1993). A neuronal group theory of sleep function. J Sleep Res, 2 (2), 63-69. Krueger, J. M., & Obal, F ., Jr. (2003). Sleep function. Front Biosci, 8 d511-519. Krueger, J. M., Obal, F., Jr., & Fang, J. (1999) Why we sleep: a theore tical view of sleep function. Sleep Med Rev, 3 (2), 119-129. Krystal, A. D., Edinger, J. D., Wohlgemuth, W. K., & Marsh, G. R. (2002). NREM sleep EEG frequency spectral correlates of sleep complaints in primary insomnia subtypes. Sleep, 25(6), 630-640. Lamarche, C. H., & Ogilvie, R. D. (1997). Elec trophysiological changes during the sleep onset period of psychophysiological insomniacs, psyc hiatric insomniacs, and normal sleepers. Sleep, 20(9), 724-733. Lavie, P. (1993). The sleep th eory of Constantin von Economo. J Sleep Res, 2 (3), 175-178. Lentz, M. J., Landis, C. A., Rothermel, J., & Shav er, J. L. (1999). Effects of selective slow wave sleep disruption on musculoskeletal pain and fatigue in middle aged women. J Rheumatol, 26(7), 1586-1592. Lichstein, K. L., & Rosenthal, T. L. (1980). Insomniacs' perceptions of cognitive versus somatic determinants of sleep disturbance. J Abnorm Psychol, 89 (1), 105-107. Lundh, L. G., & Broman, J. E. (2000). Insomnia as an interaction betwee n sleep-interfering and sleep-interpreting processes. J Psychosom Res, 49 (5), 299-310. Mahowald, M. W., & Schenck, C. H. (1992). Diss ociated states of wakefulness and sleep. Neurology, 42 (7 Suppl 6), 44-51; discussion 52. Mahowald, M. W., & Schenck, C. H. (2005). In sights from studying human sleep disorders. Nature, 437(7063), 1279-1285. Maquet, P. (2000). Functional neuroimaging of normal human sleep by positron emission tomography. J Sleep Res, 9 (3), 207-231.
88 McCall, W. V., & Edinger, J. D. (1992). Subjec tive total insomnia: an example of sleep state misperception. Sleep, 15 (1), 71-73. McCrae, C. S., Rowe, M. A., Tierney, C. G., Da utovich, N. D., Definis, A. L., & McNamara, J. P. (2005). Sleep complaints, subjective and objective sleep patterns, health, psychological adjustment, and daytime functioning in community-dwelling older adults. J Gerontol B Psychol Sci Soc Sci, 60 (4), P182-189. McCrae, C. S., Wilson, N. M., Lichstein, K. L., Durrence, H. H., Taylor, D. J., Bush, A. J., et al. (2003). 'Young old' and 'old old' poor sleeper s with and without insomnia complaints. J Psychosom Res, 54(1), 11-19. Means, M. K., Edinger, J. D., Glenn, D. M., & Fi ns, A. I. (2003). Accuracy of sleep perceptions among insomnia sufferers and normal sleepers. Sleep Med, 4 (4), 285-296. Mendelson, W. B. (1993). Pharmacologic alteration of the perception of be ing awake or asleep. Sleep, 16(7), 641-646. Mendelson, W. B. (1995a). Effect s of flurazepam and zolpidem on the perception of sleep in insomniacs. Sleep, 18(2), 92-96. Mendelson, W. B. (1995b). Effects of flurazepam and zolpidem on the perception of sleep in normal volunteers. Sleep, 18 (2), 88-91. Mendelson, W. B. (1995c). Long-term follow-up of chronic insomnia. Sleep, 18(8), 698-701. Mendelson, W. B., Garnett, D., Gillin, J. C., & Weingartner, H. (1984). The experience of insomnia and daytime and nighttime functioning. Psychiatry Res, 12 (3), 235-250. Mendelson, W. B., James, S. P., Garnett, D., Sack, D. A., & Rosenthal, N. E. (1986). A psychophysiological study of insomnia. Psychiatry Res, 19 (4), 267-284. Mendelson, W. B., & Maczaj, M. (1990). Effects of Triazolam on the Perception of Sleep and Wakefulness in Insomniacs. Annals of Clinical Psychiatry, 2 (3), 211 216. Mendelson, W. B., Martin, J. V., Stephens, H., Gi esen, H., & James, S. P. (1988). Effects of flurazepam on sleep, arousal threshold, and the perception of being asleep. Psychopharmacology (Berl), 95 (2), 258-262. Mercer, J. D., Bootzin, R. R., & Lack, L. C. ( 2002). Insomniacs' perception of wake instead of sleep. Sleep, 25 (5), 564-571. Merica, H., Blois, R., & Gaillar d, J. M. (1998). Spectral character istics of sleep EEG in chronic insomnia. Eur J Neurosci, 10 (5), 1826-1834.
89 Merica, H., & Gaillard, J. M. (1992). The EEG of the sleep onset period in insomnia: a discriminant analysis. Physiol Behav, 52 (2), 199-204. Mitchell, K. R. (1979). Behavioral treatment of presleep tension and intrusive cognitions in patients with severe predormital insomnia. J Behav Med, 2 (1), 57-69. Moldofsky, H., & Scarisbrick, P. (1976). Induction of neurasthenic musculoskeletal pain syndrome by selective sleep stage deprivation. Psychosom Med, 38 (1), 35-44. Monroe, L. J. (1967). Psychol ogical and physiological differences between good and poor sleepers. J Abnorm Psychol, 72 (3), 255-264. Morin, C. M., Blais, F., & Savard, J. (2002). Ar e changes in beliefs and attitudes about sleep related to sleep improvements in the treatment of insomnia? Behav Res Ther, 40(7), 741752. Morris, M., Lack, L., & Dawson, D. (1990). Sleeponset insomniacs have delayed temperature rhythms. Sleep, 13(1), 1-14. Mueller, J, Kiernan, R., & Langston (2001) Manual for Cognistat: The Neuobehavioral Cognitive Status Examination. Fairfax: Th e North California Neurobehavioral Group, Inc. Mukhametov, L. M., Supin, A., & Poliakova, I. G. (1984). [Sleep of Caspian seals]. Zh Vyssh Nerv Deiat Im I P Pavlova, 34 (2), 259-264. Mullaney, D. J., Kripke, D. F., & Messin, S. (198 0). Wrist-actigraphic estimation of sleep time. Sleep, 3 (1), 83-92. Nau, S. D., McCrae, C. S., Cook, K. G., & Lichst ein, K. L. (2005). Treatment of insomnia in older adults. Clin Psychol Rev, 25 (5), 645-672. Nesselroade, J. R. (2004). Intraindividual vari ability and short-term change. Commentary. Gerontology, 50(1), 44-47. Nesselroade, J. R., Mitteness, L. S., & Thom pson, L. K. (1984). Older adulthood. Short-term changes in anxiety, fatigue, a nd other psychological states. Res Aging, 6 (1), 3-23. Nesselroade, J. R., & Salthouse, T. A. (2004). Methodological and theoretical implications of intraindividual variability in perceptual-motor performance. J Gerontol B Psychol Sci Soc Sci, 59(2), P49-55. Neu, D., Mairesse, O., Hoffmann, G., Dris, A., La mbrecht, L. J., Linkowski, P., et al. (2007). Sleep quality perception in the chronic fatigue syndrome: correlations with sleep efficiency, affective symptoms and intensity of fatigue. Neuropsychobiology, 56 (1), 4046.
90 Newman, A. B., Enright, P. L., Manolio, T. A ., Haponik, E. F., & Wahl, P. W. (1997). Sleep disturbance, psychosocial correlates, and car diovascular disease in 5201 older adults: the Cardiovascular Health Study. J Am Geriatr Soc, 45 (1), 1-7. Nicassio, P. M., Mendlowitz, D. R., Fussell, J. J., & Petras, L. (1985). The phenomenology of the pre-sleep state: the developmen t of the pre-sleep arousal scale. Behav Res Ther, 23 (3), 263-271. Ohayon, M. M. (2002). Epidemiology of insomnia: what we know and what we still need to learn. Sleep Med Rev, 6 (2), 97-111. Ohayon, M. M., & Caulet, M. (1996). Psychotropic medication and insomnia complaints in two epidemiological studies. Can J Psychiatry, 41 (7), 457-464. Ohayon, M. M., Caulet, M., Priest, R. G., & Guilleminault, C. (1997). DSM-IV and ICSD-90 insomnia symptoms and sleep dissatisfaction. Br J Psychiatry, 171 382-388. Okuma, T., Matsuoka, H., Matsue, Y., & Toyomura, K. (1982). Model insomnia by methylphenidate and caffeine and use in the evaluation of temazepam. Psychopharmacology (Berl), 76 (3), 201-208. Paquet, J., Kawinska, A., & Carrier, J. (2007). Wake detection capac ity of actigraphy during sleep. Sleep, 30 (10), 1362-1369. Parrino, L., Ferrillo, F., Smerieri, A. Spaggiari, M. C., Palomba, V., Rossi, M., et al. (2004). Is insomnia a neurophysiological disorder? Th e role of sleep EEG microstructure. Brain Res Bull, 63 (5), 377-383. Perlis, M. L. (2001). Response to "Do increases in beta EEG activity uniquely reflect insomnia?" (C. H. Bastein and M. H. Bonnet). Sleep Med Rev, 5 (5), 379-383. Perlis, M. L., Giles, D. E., Bootzin, R. R., Dikm an, Z. V., Fleming, G. M., Drummond, S. P., et al. (1997). Alpha sleep and information pr ocessing, perception of sleep, pain, and arousability in fibromyalgia. Int J Neurosci, 89 (3-4), 265-280. Perlis, M. L., Giles, D. E., Mendelson, W. B., Bootzin, R. R., & Wyatt, J. K. (1997). Psychophysiological insomnia: the behavioural model and a neurocognitive perspective. J Sleep Res, 6 (3), 179-188. Perlis, M. L., Merica, H., Smith, M. T., & Giles, D. E. (2001). Beta EEG activity and insomnia. Sleep Med Rev, 5 (5), 363-374. Perlis, M. L., Smith, M. T., Andrews, P. J., Or ff, H., & Giles, D. E. (2001). Beta/Gamma EEG activity in patients with primary and sec ondary insomnia and good sleeper controls. Sleep, 24(1), 110-117.
91 Perlis, M. L., Smith, M. T., Orff, H. J., Andrew s, P. J., & Giles, D. E. (2001). The mesograde amnesia of sleep may be attenuated in subjects with primary insomnia. Physiol Behav, 74(1-2), 71-76. Perrin, F., Garcia-Larrea, L., Mauguiere, F., & Bastuji, H. (1999). A differential brain response to the subject's own name persists during sleep. Clin Neurophysiol, 110 (12), 2153-2164. Pigeon, W. R., & Perlis, M. L. (2006). Sl eep homeostasis in primary insomnia. Sleep Med Rev, 10(4), 247-254. Pilowsky, I., Crettenden, I., & Townley, M. (1985) Sleep disturbance in pain clinic patients. Pain, 23(1), 27-33. Raymond, I., Nielsen, T. A., Lavigne, G., Manzini, C., & Choiniere, M. (2001). Quality of sleep and its daily relationship to pain intensity in hospitalized adult burn patients. Pain, 92(3), 381-388. Reite, M., Buysse, D., Reynolds, C., & Mendels on, W. (1995). The use of polysomnography in the evaluation of insomnia. Sleep, 18 (1), 58-70. Reynolds, C. F., 3rd, Kupfer, D. J., Buysse, D. J., Coble, P. A., & Yeager, A. (1991). Subtyping DSM-III-R primary insomnia: a literature review by the DSM-IV Work Group on Sleep Disorders. Am J Psychiatry, 148 (4), 432-438. Riemann, D., Klein, T., Rodenbeck, A., Feige, B., Horny, A., Hummel, R., et al. (2002). Nocturnal cortisol and melatonin secretion in primary insomnia. Psychiatry Res, 113 (12), 17-27. Rioux, I., Tremblay, S., & Bastien, C. H. (2006). Time estimation in chronic insomnia sufferers. Sleep, 29(4), 486-493. Roehrs, T., Zorick, F., Koshorek, G. L., W ittig, R., & Roth, T. (1983). Effects of acute administration of brotizolam in subjects with disturbed sleep. Br J Clin Pharmacol, 16 Suppl 2, 371S-376S. Roth, C., Achermann, P., & Borbely, A. A. (1999) Frequency and state specific hemispheric asymmetries in the human sleep EEG. Neurosci Lett, 271 (3), 139-142. Saletu-Zyhlarz, G. M., Abu-Bakr, M. H., Anderer, P., Gruber, G., Mandl, M., Strobl, R., et al. (2002). Insomnia in depression: differences in objective and subjective sleep and awakening quality to normal contro ls and acute effects of trazodone. Prog Neuropsychopharmacol Biol Psychiatry, 26 (2), 249-260. Salin-Pascual, R. J., Roehrs, T. A., Merlotti, L. A., Zorick, F., & Roth, T. (1992). Long-term study of the sleep of insomnia patients with sleep state mi sperception and other insomnia patients. Am J Psychiatry, 149 (7), 904-908.
92 Saper, C. B., Chou, T. C., & Scammell, T. E. ( 2001). The sleep switch: hypothalamic control of sleep and wakefulness. Trends Neurosci, 24 (12), 726-731. Schneider-Helmert, D., & Kumar, A. (1995). Sleep, its subjective perception, and daytime performance in insomniacs with a pattern of alpha sleep. Biol Psychiatry, 37 (2), 99-105. Schwartz, G. E., Davidson, R. J., & Goleman, D. J. (1978). Patterning of cognitive and somatic processes in the self-regulati on of anxiety: effects of meditation versus exercise. Psychosom Med, 40 (4), 321-328. Semler, C. N., & Harvey, A. G. (2005). Misperce ption of sleep can adversely affect daytime functioning in insomnia. Behav Res Ther, 43(7), 843-856. Sewitch, D. E. (1984). The perceptu al uncertainty of having slept: the inability to discriminate electroencephalographic sleep from wakefulness. Psychophysiology, 21 (3), 243-259. Smith, S., & Trinder, J. (2000). Th e effect of arousals during sleep onset on estimates of sleep onset latency. J Sleep Res, 9 (2), 129-135. Stepanski, E., Zorick, F., Roehrs, T., Young, D., & Roth, T. (1988). Daytime alertness in patients with chronic insomnia compared w ith asymptomatic control subjects. Sleep, 11 (1), 54-60. Steptoe, A., & Kearsley, N. ( 1990). Cognitive and somatic anxiety. Behav Res Ther, 28 (1), 7581. Stoller, M. K. (1994). Economic effects of insomnia. Clin Ther, 16 (5), 873-897; discussion 854. Sugerman, J. L., Stern, J. A., & Walsh, J. K. (1985). Daytime alertn ess in subjective and objective insomnia: some preliminary findings. Biol Psychiatry, 20 (7), 741-750. Tang, N. K., & Harvey, A. G. (2004a). Correcting distorted perception of sleep in insomnia: a novel behavioural experiment? Behav Res Ther, 42(1), 27-39. Tang, N. K., & Harvey, A. G. (2004b). Effects of cognitive arousal and physiological arousal on sleep perception. Sleep, 27 (1), 69-78. Tang, N. K., & Harvey, A. G. (2005). Time estimati on ability and distorted perception of sleep in insomnia. Behav Sleep Med, 3 (3), 134-150. Tang, N. K., & Harvey, A. G. (2006). Altering misp erception of sleep in insomnia: behavioral experiment versus verbal feedback. J Consult Clin Psychol, 74 (4), 767-776. Taylor, D. J., Lichstein, K. L., & Durrence, H. H. (2003). Insomn ia as a health risk factor. Behav Sleep Med, 1 (4), 227-247.
93 Trajanovic, N. N., Radivojevic, V., Kaushans ky, Y., & Shapiro, C. M. (2007). Positive sleep state misperception a new c oncept of sleep misperception. Sleep Med, 8 (2), 111-118. Trinder, J. (1988). Subjective insomnia w ithout objective findings : a pseudo diagnostic classification? Psychol Bull, 103 (1), 87-94. Tryon, W. W. (2004). Issues of valid ity in actigraphic sleep assessment. Sleep, 27(1), 158-165. Ursin, H., & Eriksen, H. (2007). Cognitive activati on theory of stress, sensitization, and common health complaints. Ann N Y Acad Sci, 1113 304-310. Ursin, H., & Eriksen, H. R. (2004). The c ognitive activation theory of stress. Psychoneuroendocrinology, 29(5), 567-592. Vallieres, A., Ivers, H., Bastien, C. H., Beaulieu-Bonneau, S., & Morin, C. M. (2005). Variability and predictability in sleep patterns of ch ronic insomniacs. J Sleep Res, 14 (4), 447-453. Vallieres, A., & Morin, C. M. (2003). Actig raphy in the assessment of insomnia. Sleep, 26(7), 902-906. Van den Hoofdakker, R. H., & Beersma, D. G. (1988). On the contribution of sleep wake physiology to the explanation and the treatment of depression. Acta Psychiatr Scand Suppl, 341, 53-71. Van Egeren, L., Haynes, S. N., Franzen, M., & Hamilton, J. (1983). Presleep cognitions and attributions in sleep-onset insomnia. J Behav Med, 6 (2), 217-232. Vanable, P. A., Aikens, J. E., Tadimeti, L., Caruana-Montaldo, B., & Mendelson, W. B. (2000). Sleep latency and duration estimates among sl eep disorder patients: variability as a function of sleep disorder diagnosis, sleep hi story, and psychological characteristics. Sleep, 23(1), 71-79. Varkevisser, M., Van Dongen, H. P., & Kerkhof, G. A. (2005). Physiologic indexes in chronic insomnia during a constant routine: evidence for general hyperarousal? Sleep, 28 (12), 1588-1596. Varkevisser, M., Van Dongen, H. P., & Kerkhof, G. A. (2006). Hyperarousal as a basis for chronic insomnia: statistical misconc eptions and individual differences. Sleep, 29 (5), 719. Vgontzas, A. N., Tsigos, C., Bixler, E. O., Strataki s, C. A., Zachman, K., Kales, A., et al. (1998). Chronic insomnia and activity of the stress system: a preliminary study. J Psychosom Res, 45(1 Spec No), 21-31.
94 Vitiello, M. V., Moe, K. E., & Prinz, P. N. (2002). Sleep complaints cosegregate with illness in older adults: clinical research informed by and informing epidemiological studies of sleep. J Psychosom Res, 53 (1), 555-559. Vyazovskiy, V., Borbely, A. A., & Tobler, I. (2000). Unilateral vibris sae stimulation during waking induces interhemispheric EEG asymmetr y during subsequent sleep in the rat. J Sleep Res, 9 (4), 367-371. Vyazovskiy, V. V., Borbely, A. A., & Tobler, I. (2002). Interhemispheric sleep EEG asymmetry in the rat is enhanced by sleep deprivation. J Neurophysiol, 88 (5), 2280-2286. Walsh, J. K., & Engelhardt, C. L. (1999). The di rect economic costs of insomnia in the United States for 1995. Sleep, 22 Suppl 2 S386-393. Wehrle, R., Czisch, M., Kaufmann, C., Wetter, T. C., Holsboer, F., Auer, D. P., et al. (2005). Rapid eye movement-related brain activati on in human sleep: a functional magnetic resonance imaging study. Neuroreport, 16 (8), 853-857. Werth, E., Achermann, P., & Borbely, A. A. ( 1997). Fronto-occipital EEG power gradients in human sleep. J Sleep Res, 6 (2), 102-112. Werth, E., Dijk, D. J., Achermann, P., & Borbel y, A. A. (1996). Dynamics of the sleep EEG after an early evening nap: expe rimental data and simulations. Am J Physiol, 271 (3 Pt 2), R501-510. Wicklow, A., & Espie, C. A. (2000). Intrusiv e thoughts and their relationship to actigraphic measurement of sleep: towards a cognitive model of insomnia. Behav Res Ther, 38(7), 679-693. Wilson, K. G., Eriksson, M. Y., D'Eon, J. L., Mikail, S. F., & Emery, P. C. (2002). Major depression and insomnia in chronic pain. Clin J Pain, 18(2), 77-83. Wohlgemuth, W. K., Edinger, J. D., Fins, A. I ., & Sullivan, R. J., Jr. (1999). How many nights are enough? The short-term st ability of sleep parameters in elderly insomniacs and normal sleepers. Psychophysiology, 36 (2), 233-244. Yasuda, T., Yasuda, K., Brown, R. A., & Krueger, J. M. (2005). State-dependent effects of lightdark cycle on somatosensory and visual cortex EEG in rats. Am J Physiol Regul Integr Comp Physiol, 289 (4), R1083-1089. Yoshida, H., Peterfi, Z., Garcia-Garcia, F., Kirk patrick, R., Yasuda, T., & Krueger, J. M. (2004). State-specific asymmetries in EEG slow wa ve activity induced by local application of TNFalpha. Brain Res, 1009 (1-2), 129-136.
95 BIOGRAPHICAL SKETCH Mr. Kay is a graduate student in the Departm ent of Clinical and Health Psychology at the University of Florida studying to be a sleep specialist as a clin ical psychologist. Mr. Kay has been involved in sleep research for 5 year. His tr aining began in the animal sleep research lab of Dr. James Krueger at Washington State Universi ty. He continued his research at Brigham Young University studying the biolog ical and behavioral correlates of infants diurnal sleep in Rhesus Macaques and received the American A cademy of Sleep Medicine Young Investigator Honorable Mention Award in 2008 for his work in this area. He is currently studying sleep misperception in older adult in Dr. Christina McCr aes sleep research lab. He won the College of Public Health and Health Profession (PHHP ) Young Research Award and the Department of Clinical and Health Psychology Best Fall Sympos ium Presentation Award in 2008 on his work in the area of sleep misperception. Specific res earch interests include the functions of sleep, localization of brain functions during sleep, sleep in development, and the sleeping brain in mental and physical illnesses. Mr. Kay is committed to helping make cognitive-behavioral treatments for sleep problems mo re integrative, accessible, and e ffective in clinical practice.