<%BANNER%>

Actigraphy and Sleep Diaries in the Assessment of Sleep Patterns in Cardiac Disease

Permanent Link: http://ufdc.ufl.edu/UFE0024055/00001

Material Information

Title: Actigraphy and Sleep Diaries in the Assessment of Sleep Patterns in Cardiac Disease
Physical Description: 1 online resource (95 p.)
Language: english
Creator: Cross, Natalie
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: actigraphy, cad, cardiac, diary, icd, sleep
Clinical and Health Psychology -- Dissertations, Academic -- UF
Genre: Psychology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Cardiac patients free from obstructive sleep apnea (OSA) frequently have insomnia symptoms that pose additional risk for future cardiac events. Poor sleep may relate to somatic and psychological hyperarousal that leads to short total sleep time and poor sleep quality. Women who sleep less than or equal to six hours or greater than or equal to nine hours per night are significantly more likely to experience myocardial infarction in the 10 years subsequent to these sleep patterns than women who sleep 7-8 hours per night (Ayas et al., 2003). We hypothesized that implantable cardioverter defibrillator (ICD) patients will have poorer sleep than coronary artery disease (CAD) patients related to hypervigilance for device functioning and worry over shock discharge. No studies to date have used objective measures of sleep to compare sleep patterns among CAD and ICD patients. We investigated sleep efficiency (a percentage of time spent sleeping divided by time spent in bed) and sleep latency (time between lights out and sleep initiation) in a sample of 60 patients (n = 30 CAD and n = 30 ICD) without OSA at the University of Florida & Shands Hospital. For 14 days, participants completed a daily sleep diary. Additionally, half of the total sample also used actigraphy for 14 days to objectively measure their sleep. An actigraph is watch-like device that infers wakefulness or sleep from the presence or absence of limb movements. Actigraphy strongly correlates with polysomnographic measures of sleep. Using actigraphy, mean sleep efficiency was poorer (69.76%) in CAD patients and more adaptive (82.80%) in ICD patients. This difference was highly significant, F(1,27) = 16.840, p < .001. CAD patients also had shorter mean total sleep times per sleep diaries (336.19 minutes; 5.60 hours) compared to ICD patients (430.65 minutes; 7.18 hours), F(1,27) = 15.908, p < .001. Ejection fraction% (EF%) and physical activity significantly predicted 56% of the variance in sleep efficiency scores, F(2,26) = 8.322, p = .005. The finding that ICD patients slept more efficiently than CAD patients is surprising given that CAD patients had higher EF%s. This difference cannot be accounted for by differences in somatic hypervigilance, depression, anxiety, or physical activity levels. Results suggest that chest pain in patients with CAD may be an important consideration in analyzing factors that influence sleep in cardiac patients. Further research is needed to further stratify the exact contribution of both medical and psychological factors in CAD patients.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Natalie Cross.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Sears, Samuel F.
Local: Co-adviser: McCrae, Christina S.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2011-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024055:00001

Permanent Link: http://ufdc.ufl.edu/UFE0024055/00001

Material Information

Title: Actigraphy and Sleep Diaries in the Assessment of Sleep Patterns in Cardiac Disease
Physical Description: 1 online resource (95 p.)
Language: english
Creator: Cross, Natalie
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: actigraphy, cad, cardiac, diary, icd, sleep
Clinical and Health Psychology -- Dissertations, Academic -- UF
Genre: Psychology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Cardiac patients free from obstructive sleep apnea (OSA) frequently have insomnia symptoms that pose additional risk for future cardiac events. Poor sleep may relate to somatic and psychological hyperarousal that leads to short total sleep time and poor sleep quality. Women who sleep less than or equal to six hours or greater than or equal to nine hours per night are significantly more likely to experience myocardial infarction in the 10 years subsequent to these sleep patterns than women who sleep 7-8 hours per night (Ayas et al., 2003). We hypothesized that implantable cardioverter defibrillator (ICD) patients will have poorer sleep than coronary artery disease (CAD) patients related to hypervigilance for device functioning and worry over shock discharge. No studies to date have used objective measures of sleep to compare sleep patterns among CAD and ICD patients. We investigated sleep efficiency (a percentage of time spent sleeping divided by time spent in bed) and sleep latency (time between lights out and sleep initiation) in a sample of 60 patients (n = 30 CAD and n = 30 ICD) without OSA at the University of Florida & Shands Hospital. For 14 days, participants completed a daily sleep diary. Additionally, half of the total sample also used actigraphy for 14 days to objectively measure their sleep. An actigraph is watch-like device that infers wakefulness or sleep from the presence or absence of limb movements. Actigraphy strongly correlates with polysomnographic measures of sleep. Using actigraphy, mean sleep efficiency was poorer (69.76%) in CAD patients and more adaptive (82.80%) in ICD patients. This difference was highly significant, F(1,27) = 16.840, p < .001. CAD patients also had shorter mean total sleep times per sleep diaries (336.19 minutes; 5.60 hours) compared to ICD patients (430.65 minutes; 7.18 hours), F(1,27) = 15.908, p < .001. Ejection fraction% (EF%) and physical activity significantly predicted 56% of the variance in sleep efficiency scores, F(2,26) = 8.322, p = .005. The finding that ICD patients slept more efficiently than CAD patients is surprising given that CAD patients had higher EF%s. This difference cannot be accounted for by differences in somatic hypervigilance, depression, anxiety, or physical activity levels. Results suggest that chest pain in patients with CAD may be an important consideration in analyzing factors that influence sleep in cardiac patients. Further research is needed to further stratify the exact contribution of both medical and psychological factors in CAD patients.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Natalie Cross.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Sears, Samuel F.
Local: Co-adviser: McCrae, Christina S.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2011-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024055:00001


This item has the following downloads:


Full Text

PAGE 1

ACTIGRAPHY AND SLEEP DIAR IES IN THE ASSESSMENT OF SLEEP PATTERNS IN CARDIAC DISEASE By NATALIE JOAN CROSS A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2009 1

PAGE 2

2009 Natalie Joan Cross 2

PAGE 3

To Chris, my husband and best friend in the whole wide world; and to Eli, who keeps my eye on whats most important in my world. 3

PAGE 4

ACKNOWLEDGMENTS Obtaining a doctorate degree has been an honor a true luxury, and a grueling task all in one. I could have never made it this far if it hadnt been for my parents, Patrick William and Cristine Marie Small. Their fortitude and persev erance in obtaining their own degrees in higher education have been a large part of my inspirat ion to pursue a doctorate in my chosen field. Their focus on education and cultiv ation of curiosity in raising me made me who I am today. My sisters Vanessa and Audrey have given me c onstant support and reminded me to have some fun in life. Im extremely grateful to my chair, Samuel Sears, Ph.D., for his mentorship on this study and throughout my years in graduate school. Acro ss research projects, co ursework, and practica, he has brought a wide range of knowledge, a sense of individuality, c onfidence, and a strong presence to our interactions. Whether a Gator or a Pirate, Dr. Sears level of passion for his research and clinical work with cardiac patients has been both admirable and inspirational. Im also lucky to have an excellent doct oral committee that shepherded me through the process of obtaining my doctoral degree. My co-chair, Christina McCrae, Ph.D., opened a world of interest for me by teaching me about behavioral sleep medicine. Her skillful and consistent supervision allowed me to learn a great deal about this population and to feel confident about my decision to work with sleep patients in the future I thank Deidre Pereira, Ph.D., David Janicke, Ph.D., and James Jessup, Ph.D. as committee members, as they have been informative and supportive. Big thanks are in order to Neha Dixit. Her sincere friendship and emotional support throughout this project has been a true lifesaver. She has alwa ys been there for a laugh or commiseration, whichever was needed. 4

PAGE 5

Most importantly, I am utterly grateful to my husband, Christopher Wayne Cross. Cheers to a man whos given me unconditi onal love, encouragement, and support during these years. I aspire to be the kind of partner that he has been to me. Chris sense of humor has helped me make it through the ups and downs of graduate school. His willingness to play Mr. Mom during the past few months has been key to comple ting this project. Last and anything but least is Elijah Cross. I thank Eli for making me laugh and bringing me so much joy! 5

PAGE 6

TABLE OF CONTENTS page ACKNOWLEDGMENTS ...............................................................................................................4 LIST OF TABLES ...........................................................................................................................8 LIST OF FIGURES .......................................................................................................................10 ABSTRACT ...................................................................................................................................11 CHAPTER 1 INTRODUCTION................................................................................................................. .13 Review of the Literature .........................................................................................................14 Overview of Sleep Disturbances in Cardiac Disease ......................................................14 Hypertension and Sleep ...................................................................................................15 Coronary Artery Disease and Sleep................................................................................16 Cardiac Arrhythmias, Implantable Card ioverter Defibrillators, and Sleep .....................18 Comorbid Insomnia .........................................................................................................20 Insomnia in Older Adults ................................................................................................22 Psychological Distress, Cardiac Disease, and Sleep .......................................................24 Actigraphy Studies on Sleep in Chronic Disease ............................................................28 Summary and Hypotheses Regardi ng Sleep and Cardiac Disease .........................................30 2 METHODS...................................................................................................................... .......32 Participants and Recruitment ..................................................................................................32 Measures .................................................................................................................................33 Demographic Information ...............................................................................................33 Biomedical Data ..............................................................................................................33 Objective Sleep Data .......................................................................................................34 Subjective Sleep Data ......................................................................................................35 Procedures ...............................................................................................................................38 Statistical Analyses ..........................................................................................................41 Power and Sample Size Calculations ..............................................................................43 3 RESULTS...................................................................................................................... .........44 Demographic and Medical Variable Desc riptives and Cardiac Group Differences...............44 Psychosocial and Sleep Variable Desc riptives and Cardiac Group Differences ....................45 Aim 1: Assessment of ICD and CAD Group Differences for Sleep Diary and Actigraphy Data ..................................................................................................................47 Aim 2: Prediction Model.......................................................................................................49 Hierarchical Regression Models ......................................................................................50 Multiple Regression Models ............................................................................................51 6

PAGE 7

4 DISCUSSION................................................................................................................... ......77 Patient Characteristics ............................................................................................................77 CAD and ICD Participant Differences in Sleep .....................................................................80 Predictors of Sleep Efficiency in the Full, CAD, and ICD Samples ......................................82 Limitations ..............................................................................................................................83 Significance ............................................................................................................................84 Conclusions .............................................................................................................................85 LIST OF REFERENCES ...............................................................................................................86 BIOGRAPHICAL SKETCH .........................................................................................................95 7

PAGE 8

LIST OF TABLES Table page 1-1 Commonly Used Sleep Indices (Speci fic to Sleep Diar y Data Collection) .......................31 2-1 Psychosocial Measures ......................................................................................................43 3-1 Descriptive statistics on demographic, medical, sleep, and psychosocial variables for overall sample ....................................................................................................................54 3-2 Means and standard deviations/frequencies for medical, sleep, and psychosocial variables by cardiac condition ...........................................................................................56 3-3 Means and standard deviations for sleep variables for sleep diary study arm and actigraphy data from sleep diary plus actigraphy arm .......................................................57 3-4 Pearsons product-moment correl ations among psychosocial variables ...........................58 3-5 Pearsonss product-moment correlations among sleep variables......................................59 3-6 Pearsons product-moment correlations among sleep efficiency and corresponding predictors for full sample of actigraphy indices ................................................................60 3-7 Pearsons product-moment correlations among sleep efficiency and corresponding predictors for full sample of sleep diary indices ................................................................61 3-8 Pearsons product-moment correlations among sleep efficiency and corresponding predictors for CAD samp le of actigraphy indices ..............................................................62 3-9 Pearsons product-moment correlations among sleep efficiency and corresponding predictors for CAD sample of sleep diary indices .............................................................63 3-10 Pearsons product-moment correlations among sleep efficiency and corresponding predictors for ICD sample of actigraphy indices ...............................................................64 3-11 Pearsons product-moment correlations among sleep efficiency and corresponding predictors for ICD sample of sleep diary indices ...............................................................65 3-12 ANCOVA results for mean sleep effici ency (sleep diary) by cardiac condition ...............66 3-13 ANCOVA results for mean number of awakenings (sleep diary) by cardiac condition ...66 3-14 ANCOVA results for mean sleep qual ity (sleep diary) by cardiac condition ....................67 3-15 ANCOVA results for mean sleep onse t latency (actigraphy) by cardiac condition ..........67 3-17 ANOVA results for mean sleep efficiency (actigraphy) by cardiac condition ..................68 8

PAGE 9

3-18 ANOVA results for mean total sleep time (actigraphy) by cardiac condition. ..................69 3-19 ANOVA results for mean waking after slee p onset (actigraphy) by cardiac condition. ...69 3-20 ANOVA results for mean total sleep time (sleep diary) by cardiac condition ..................70 3-21 Kruskal-Wallis ANOVA results for mean sleep onset latency (sleep diary) by cardiac condition ............................................................................................................................70 3-22 Kruskal-Wallis ANOVA results for mean waking after sleep onset (sleep diary) by cardiac condition ................................................................................................................70 3-23 Summary of hierarchical regression analysis for predic tors of sleep diary sleep efficiency for overall sample .............................................................................................71 3-24 Summary of hierarchical regression analysis for pr edictors of actigraphy sleep efficiency for overall sample .............................................................................................72 3-25 Summary of hierarchical regression analysis for predic tors of sleep diary sleep efficiency for CAD sample ................................................................................................73 3-26 Summary of hierarchical regression analysis for predic tors of sleep diary sleep efficiency using ICD sample..............................................................................................74 3-27 Summary of hierarchical regression analysis for pr edictors of actigraphy sleep efficiency using CAD sample............................................................................................75 3-28 Summary of multiple regression analysis fo r predictors of actigraphy sleep efficiency using full sample ................................................................................................................76 3-29 Summary of multiple regression analysis for predictors of sleep diary sleep efficiency using ICD sample..............................................................................................76 9

PAGE 10

LIST OF FIGURES Figure page 3-1 Sample derivation and data collection process ..................................................................53 10

PAGE 11

Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy ACTIGRAPHY AND SLEEP DIAR IES IN THE ASSESSMENT OF SLEEP PATTERNS IN CARDIAC DISEASE By Natalie Joan Cross August 2009 Chair: Samuel F. Sears Cochair: Christina S. McCrae Major: Psychology Cardiac patients free from obstructive sleep apnea (OSA) frequently have insomnia symptoms that pose additional risk for future card iac events. Poor sleep may relate to somatic and psychological hyperarousal th at leads to short total sleep time and poor sleep quality. Women who sleep less than or equal to six hours or greater than or equal to nine hours per night are significantly more likely to experience myoc ardial infarction in the 10 years subsequent to these sleep patterns than women who sleep 78 hours per night (Ayas et al., 2003). We hypothesized that implantable cardioverter defibr illator (ICD) patients will have poorer sleep than coronary artery disease (CAD) patients re lated to hypervigilance fo r device functioning and worry over shock discharge. No studies to da te have used objective measures of sleep to compare sleep patterns among CAD and ICD patients. We investigated sleep efficiency (a per centage of time spent sleeping divided by time spent in bed) and sleep latency (time between ligh ts out and sleep initiation) in a sample of 60 patients (n = 30 CAD and n = 30 ICD) without OSA at the Un iversity of Florida & Shands Hospital. For 14 days, participants completed a da ily sleep diary. Additiona lly, half of the total sample also used actigraphy for 14 days to objectively measure their sleep. An actigraph is 11

PAGE 12

watch-like device that infers wakefulness or sleep from the presence or absence of limb movements. Actigraphy strongly correlates with polysomnographic measures of sleep. Using actigraphy, mean sleep efficiency was poorer (69.76%) in CAD patients and more adaptive (82.80%) in ICD patients. Th is difference was highly significant, F (1,27) = 16.840, p < .001. CAD patients also had shorter mean tota l sleep times per sleep diaries (336.19 minutes; 5.60 hours) compared to ICD pa tients (430.65 minutes; 7.18 hours), F (1,27) = 15.908, p < .001. Ejection fraction% (EF%) and physi cal activity significantly predic ted 56% of the variance in sleep efficiency scores, F (2,26) = 8.322, p = .005. The finding that ICD patients sl ept more efficiently than CAD patients is surprising given that CAD patients had higher EF%s. This diffe rence cannot be accounted for by differences in somatic hypervigilance, depression, anxiety, or physical activity le vels. Results suggest that chest pain in patients with CAD may be an impor tant consideration in analyzing factors that influence sleep in cardiac patients. Further research is needed to further stratify the exact contribution of both medi cal and psychological fact ors in CAD patients. 12

PAGE 13

CHAPTER 1 INTRODUCTION Initial evidence examining sleep in cardiac patients has implicated sleep patterns as potentially contributory to key health outcomes (Ayas et al., 2003; Gangwisch et al., 2006; McSweeney et al., 2003). Cardiac patients with maladaptive sleep patterns experience additional risk for future cardiac events due to factors in cluding inappropriate total sleep time and poor sleep quality. Women who sleep < 6 hours or > 9 hours per night are significantly more likely to experience MI in the 10 years subsequent to these sleep patt erns than women who sleep 7-8 hours per night (Ayas et al., 2003) Women who experience myo cardial infarction (MI) report that sleep disturbance is the second most co mmonly reported symptom in the month preceding the MI (McSweeney et al., 2003). These studies are some of the first to examine sleep patterns in cardiac disease and provide a basis for unde rstanding how maladaptive sleep patterns can present an additional risk for cardiac events ove r the risk that is inherent to normal sleep. The current study was designed to assess the feasibility of us ing sleep diaries as a method of measuring standard sleep indices (such as sleep efficiency, total sleep time, and waking after sleep onset) in cardiac patients across a period of two weeks. While most of the previous research on sleep in cardiac disease has focused on patients with obstructive sleep apnea, this study used coronary artery disease (CAD) and implantable cardioverte r defibrillator (ICD) patients. Objective sleep data from actigraphy wa s used as a comparison method in a subset of patients to test for correspondence with subjectiv e sleep data from standard two-week sleep diaries. Specifically, this study yields important inform ation as to whether sleep efficiency, total sleep time, sleep onset latency, and waking af ter sleep onset differ between CAD and ICD patients with respect to both objective and subjective data using Independent Samples t -tests. It 13

PAGE 14

was hypothesized that ICD patient s would have poorer sleep i ndex values than CAD patients related to hypervigilance for device functioning a nd worry over nocturnal shock discharge. The study also examined predictors of sleep efficien cy in the overall study sample patients using a hierarchical regression analysis. It was expected that biomarkers (age, sex, and left ventricular ejection fraction (EF%), psychological distre ss, somatic hypervigilance, daytime physical activity levels, and other standa rd sleep indices would serve as significant predictors. Results from this study produced valuable information by documenting the feasibility of using sleep diaries as a data co llection method in cardiac patients While a small collection of studies have previously documented the utility of using actigraphy in CAD patients, this study will add to the literature by providing prelim inary information on the feasibility of using actigraphy with electrophysiology (ICD) patients in particular. Review of the Literature Overview of Sleep Disturbances in Cardiac Disease In the management of cardiac disease, primary to tertiary care has e ndeavored to intervene on all key areas of the disease. Sleep quality may be a valuable and under-examined variable that is currently gaining attention in the rese arch literature. Incr eased attention on sleep problems has come about at a time when American s are more likely to die from cardiac disease than any other health condition (AHA, 2006). Also, Americans now sleep fewer hours per night than ever before 1 (NSF, 2005). Clinical psychologists inte rvene in a target-rich environment. The management of behavioral risk factors for cardiovascular disease has traditionally entailed reducing the detrimental effects of hypertension, c holesterol, tobacco use, psychosocial stress, 1 Vorona (2005) discussed the discrepancy in sleep patterns from 1900 to 2000. In 2005, the average American slept 7.1 hours per night. This is much lower than the length of time that the average Ameri can slept per night 100 years ago, approximately 9 hours. 14

PAGE 15

poor dietary habits, and a sedentary lifestyle on the heart. In addition to these behaviors, poor sleep is a prevalent and clinical ly significant problem that cardiac psychologists can effectively treat. Although sleep improve ments can increase quality of life and decrease mortality (Piccirillo, Duntley, & Schotland, 2000; Verb eek, Konings, Aldenkamp, Declerck, & Klip, 2006), sleep has largely remained a potential silent contributor to disease until recently. Hypertension and Sleep Hypertension is a well-examined risk factor for the initiation and progression of cardiac disease. The prevalence of hypertension has in creased during the past 10 years, notwithstanding improvements in awareness, treatment, and contro l of the disease (Gangwisch et al., 2006). Hypertension is one mechanism that explains th e relationship between sl eep disturbance and the incidence of MI. Sleep provide s an opportunity for normative decr eases of 10 to 20% in average 24-hour blood pressure. Patients who sleep fewer hours per night experience the same proportional decrease in blood pressure while sleeping, although thei r average daily blood pressure increases as a result of being awake for more hours per day (Gangwisch et al., 2006). Decreased time spent sleeping a nd therefore increased waking hour s often means that there is more time for exposure to elevated sympathetic nervous system activity, for eating, and for exposure to psychosocial stressors, all of which po se additional risk for coronary artery disease (Vorona et al., 2005). As part of the National Health and Nutr ition Examination Survey I (NHANES I; Gangwisch et al., 2006), researchers examined se lf-reported sleep duration and the incidence of new hypertension in a sample of 4810 participan ts. Sleep duration was reported during the baseline data collection from 1982-1984, while h ypertension incidence was determined from follow-up data gathered in 1986, 1987, and 1992. Resu lts indicated that participants with both short sleep duration ( < 5 hours per night) and long sleep duration ( > 9 hours per night) were 15

PAGE 16

significantly more likely to have a diagnosis of hypertension compar ed to participants who slept 7-8 hours nightly. For the short sleep duration group, these results held even when several demographic and health behavior variables were covaried, including daytime sleepiness, depression, physical activity, alcohol consumption, salt consump tion, smoking, pulse rate, and gender. Analysis by age revealed that it was pr imarily short sleep duration that was associated with hypertension in young participants, while it was primarily long sleep duration that was associated with hypertension in older participan ts. These age differences may have occurred because short sleep duration is more strongly associated with obesity in younger versus older participants (Gangwisch et al ., 2006). The NHANES I study resu lts suggest that short sleep duration may contribute to the e tiology of hypertension in some pa tients, and interventions that serve to improve total sleep time may aid in the primary prevention of hypertension. A lack of symptoms and warning signs for hypertension allows approximately 30% of hypertensive individuals to not realize their condition (Z usman, 2001), although the condition is highly prevalent and acts as a global mechanism linking sleep problems with cardiac disease. The following two sections summarize relations between sleep patterns and the tw o groups of cardiac patients under study, CAD and ICD patients. Coronary Artery Disease and Sleep In the context of CAD, sleep duration a nd sleep quality appear to have important implications for morbidity and mortality. For example, Ayas and colleagues (2003) examined total sleep time as a risk factor for coronary events in over 71,000 women as part of the Nurses Health Study. The primary outcome measure, 10-year incidence for coronary events, was defined as a nonfatal MI. Partic ipants provided information on thei r total hours of actual sleep in a 24-hour period in 1986. Among healthy participants free of cardiac disease at study enrollment, the rates of prospective coronary events were measured from 1986 to 1996. Results 16

PAGE 17

indicated that both short total sleep times and long total sleep times were associated with a risk for future cardiac problems. On the average, those who slept < 5 or 6 hours nightly had a relative risk of 1.45 or 1.18, respec tively, for future MI. Similarly, women who slept 9 hours had a relative risk of 1.36 for future MI. These results were found when pertinent medical and lifestyle factors were covari ed, including hypertension, body ma ss index (BMI), lipids, and tobacco use (Ayas et al, 2003). Sleep quality is a second key construct that has been examined as a correlate of cardiac disease. McSweeney and colleagues (2003) stud ied sleep quality in women who had suffered from MI within the six months before study en try. Patients were asked retrospectively about their prodromal and acute physical symptoms in the month preceding their MI, and 54% of participants endorsed sleep distur bance (defined as poor subjectiv e sleep quality) as a prodromal symptom. Sleep disturbance was the second mo st frequently reported symptom and was second only to fatigue, a construct that frequently rela tes to poor subjective sleep (Lavidor, Weller, & Babkoff, 2003). Interestingly, sleep disturbance was more frequen tly cited as a health concern than commonly cited cardiac concerns such as ches t or neck pain. Of note, a full half of women who reported sleep problems before MI rated the problems as sev ere in intensity (McSweeney et al., 2003). Taken together, the Ayas (2003) and McSweeney (2003) studies provide strong preliminary evidence for sleep as a correlate of cardiac disease outcomes in patients who experience MI. However, the lite rature is lacking studies that have examined a more complete, holistic array of sleep indices in CAD patients, including total sleep time, sleep quality, sleep efficiency, sleep onset latency, and waking after sleep onset. Such research could increase our understanding of how CAD patients sleep and potentially present a ppropriate areas for psychological and sleep-related interventions. 17

PAGE 18

Cardiac Arrhythmias, Implantable Ca rdioverter Defibrillators, and Sleep Slow-wave sleep, which accounts for 20% of total sleep time, is usually a time of increased baroreceptor sensitivity, increa sed parasympathetic tone, and decreased heart rate (Lavery, Mittleman, Cohen, Muller, & Verrier, 1997). These non-REM (non-rapid eye movement) periods of slow-wave sleep are when blood pre ssures reach their lowest point of the 24-hour period. While slow-wave sleep is generally tho ught of as a time of repair and restoration, REM sleep, which accounts for the remaining 20% of total sleep time, is a time of decreased respiratory control and vulnerability to surges in autonomic activity that can contribute to cardiac arrhythmias (Verrier & Mittleman, 2005; Verri er & Josephson, 2005). Specifically, these increases in autonomic arousal can lead to surges or pauses in heart rhyt hm, increased electrical instability, and ventricular tachyarrhythmias (Lav ery et al., 1997). Su rges in blood pressure, heart rate, and strong or violent emotions related to dreaming have been associated with the occurrence of MI or sudden cardiac death during REM (Verrier & Mittlemen, 2005). Nocturnal arrhythmogenesis is possible for patients with normal blood pressure because normal sleep patterns contain these autonomic surges in activity that are an alogous to the autonomic surges experienced during waking hours wi th physical exertion and sexua l activity. Hypertentensive patients are especially at risk for arrhythm ic activity because normative REM surges in autonomic activity can superimpose on blood pressure s that are already considered hypertensive (Verrier & Josephson, 2005). Lavery et al. (1997) analyzed data acro ss 12 published studies detailing the circadian pattern of sudden cardiac death. The occurr ence of sudden cardiac death was documented by device interrogation for ICD participants and by mu ltiple standard methods for other participants (e.g., arrival time of rescue squad, receipt of emergency call by dispatc h, telephone interviews with significant others, or deat h certificate). Findings indicated that approximately 15% percent 18

PAGE 19

of malignant ventricular fibrillation episode s (across both ICD and non-ICD groups) occurred during sleep between the hours of midnight and 5:59 A.M. (Lavery et al., 1997). The most common form of arrhyt hmia, atrial fibrillation (AF), has been shown to have a higher incidence from midnight to 5:00 AM than at other points in the circadian rhythm (Verrier & Josephson, 2005), and this has been termed vaga lly mediated AF. Approximately 30% of AF episodes occur while sleeping, and the prevalence of AF increases significantly when it cooccurs with obstructive sleep apnea, where hypoxi a and hemodynamic stress are believed to be most problematic. Kanagala et al. (2003) demonstrated that pa tients with untreated obstructive sleep apnea (OSA) have a higher risk of AF recurrent successf ul cardioversion than patients without documented OSA. Additionally, OSA patie nts treated with continuous positive airway pressure (CPAP) had a recurrence rate of 42%, whereas untreated or inappropriately treated OSA patients experienced AF at 82% (Kangala et al ., 2003), suggesting that th e incidence for some arrhythmias can be decreased by treating so me forms of sleep-disordered breathing appropriately. In a seminal study, Guilleminault and colleagues (1983) measured arrhythmias in 400 patients with OSA using 24-hour Holter mon itoring and found that nearly half (48%) experienced cardiac arrhythmias dur ing one given night of sleep. Of those who had arrhythmic activity, 2% had sustained ventri cular tachycardia, 11% had sinus arrest, 8% had second-degree atrioventricular block, and 19% had premature ve ntricular contractions. Therefore, sleeping periods may have intrinsic value in unde rstanding the genesis of arrhythmias. Two studies at the University of Florida have examined patients with ICDs and found that they experience sleep problems at relatively high rates that impact quality of life. Serber and colleagues (2003) reported that 87 % of Implantable Atrial Defibrillation patients had clinically 19

PAGE 20

significant sleep difficulties as evidenced by results on the Pittsburgh Sleep Quality Index (PSQI). There were no group differences fo r sleep in shocked versus nonshocked patients (Serber et al., 2003). Similarly, Cross and colleagues (2006) surveyed a heterogeneous sample of 69 cardiac patients, and a subset of this group were ICD patients ( n = 18). In this small sample, the mean PSQI score was 8.07 (where scores of 6 and higher indicate clinically significant poor sleep quality) a nd 57.1% of the sample demonstrat ed clinically significant sleep problems. Poor sleep may occur in ICD patients for a variety of reasons, including a history of nocturnal shocks, depression, noctu rnal angina or chest pain, pa lpitations, decreased physical activity, orthopnea, sleep-disorder ed breathing, or pharmacologic si de effects (e.g, use of betablockers; Haffajee et al., 2002; La ttimore et al., 2003). Although th e studies discussed in this section have examined how sleep patterns and RE M activity relate to patients with arrhythmias and ICDs, most of the research in this area ha s focused on the relationship between arrhythmias and obstructive sleep apnea. Studies are needed that present a clear picture of sleep patterns in arrhythmia and ICD patients by using both subjec tive and objective measures of standard sleep indices. Additionally, little in formation is currently known regarding how sleep patterns may differ in these patient groups as compared to patients with CAD. Comorbid Insomnia Insomnia complaints can be symptom or syndrome, and a wide range of insomnia types exist. Insomnia is a generic term for poor sleep that can refe r to sleep onset insomnia, sleep maintenance insomnia, terminal insomnia, a nd/or nonrestorative sleep. See Table 1-1 for definitions of commonly used sleep indices in as sessing insomnia (as defined in sleep diary data collection). Most cases of inso mnia included difficulties with both long sleep onset latency and poor sleep maintenance (Berry, 2003). Increasi ngly, insomnia is being conceptualized as a condition involving altered physiolo gy throughout the duration of th e 24-hour day. For example, 20

PAGE 21

patients with insomnia often display increased alertness during the daytime and not just at nighttime while trying to sleep, co mpared to patients with adap tive sleep patterns (Varkevisser, Van Dongen, & Kerkhof, 2005). Spielmans 3Ps model (Spielman & Glovins ky, 1991) was used to conceptualize the relationship between cardiac dis ease and sleep problems. The 3Ps model concerns predisposing, precipitating, and perpetuating factors. This mo del asserts that predispos ing conditions such as age, gender, or level of arousability lower the th reshold needed for triggering sleep disturbance. Precipitating circumstances are th e temporal or contextual f actors (e.g., initiation of cardiac disease) that surround the actual onset of the sleep disturbanc e. Perpetuating factors are variables that contribute to the maintenance of sleep problems over time, such as maladaptive sleep habits. There are three classification systems used to diagnose and code sleep disorders: (1) The Diagnostic and Statistical Manual of Mental Disorders, Four th Edition (DSM-IV), (2) The International Classification of Disease-10 th Edition (ICD-10), and (3) The International Classification of Sleep Disorders Revised (ICS D-III). Whereas the DSM-IV and the ICSD-R distinguish between the two main types of insomnia (primary and comorbid insomnia), the ICD10 does not. Additionally, the ICSD-III gives the most detailed classification of sleep disorders. While many definitions of insomnia exist, resear chers and clinicians generally consider insomnia complaints to be of clinical significance if they involve > 31 minutes of sleep latency or waking after sleep onset for > 3 nights weekly for > 6 months (Lichstein, Durrence, Taylor, Bush, & Riedel, 2003). Primary insomnia is a condition of poor sleep that is not attributable to medical or psychiatric symptoms (Ranjan, 2005). Primary inso mnia often begins in the context of a major 21

PAGE 22

life event or stressor that leads to conditioning fo r poor sleep (Morin et al ., 2004). Subsequently, the sleep problems remain even after the precip itant has passed. Cognitive-Behavioral Therapy for Insomnia (CBT-I) is a multicomponent inte rvention that addresses sleep-incompatible behaviors, cognitions, and physiologi cal arousal. Most of the litera ture that has examined CBT-I has been in primary insomnia samples. This fact is ironic in light of the fact that 75% of insomnia cases represent comorbid insomnia (Lichstein, Nau, McCrae, & Cook, 2005). Comorbid insomnia is a condition of poor sleep that is accompanied by a medical or psychological condition (Rybarczyk et al., 2005). Complaints of comorbid insomnia relate reciprocally to medical conditions such as cardiac di sease, depression, anxi ety, substance abuse, or other sleep disorders. Id entifying which conditions are causal and which conditions are effects in the list of possible comorbid conditions for a patient is often difficult or impossible. Therefore, the key with interveni ng in comorbid insomnia is to tr eat the insomnia directly, as its own condition (Rybarczyk et al., 2005; Stepansk i & Rybarczyk, 2006). Over time and with conditioning to poor sleep patterns, insomn ia symptoms tend to become functionally autonomous; they take on a life of their own independent from the medical or psychological condition. Recent intervention rese arch with a variety of samp les, including chronic pain, nonmetastatic breast cancer, and mixed psychiatric disorder patients, has provided evidence that significant sleep improvements can occur in the context of comorbid me dical and psychological conditions (Currie, Wilson, Pontefract, & deLa plante, 2000; Lichstein et al., 2000; Quesnel, Savard, Simard, Ivers, & Morin, 2003). Insomnia in Older Adults While cardiac disease affects patients in ev ery age group, most cardiac patients are older adults and the modal age range is 55-75 (AHA, 2006). Insomnia is often worse among older adults due to (a) the effects that normative aging exerts on sleep patterns and (b) the interactive 22

PAGE 23

effects of age and chronic illness on sleep. With increasing age, changes in sleep architecture lead to decreased depth, durati on, and continuity of sleep (Mor gan, 2000). Sleep architecture is the proportion of sleep spent in the various NREM (4 stages) and REM stages of sleep. Stage 1 sleep represents a transition from wakefulness to sleep and accounts for approximately 5% of total sleep time, while Stage 2 sleep serves as a transition from light to deep sleep and comprises 50% of time spent sleeping. Stages 3 and 4 are th e deepest levels of sl eep known as slow wave sleep. These two stages combined account for 20 % of total sleep time. REM, or paradoxical sleep, comprises 25% of time spent sleeping and is followed by repetitions of this cycle (Morgan, 2000). One sleep cycle lasts 90-100 minutes (Lacks & Morin, 1992) such that approximately five cycles occur during a standa rd eight hours of slee p. Although there is not consensus on this, experts genera lly agree that slow wave sleep is a time of somatic restoration and REM sleep is a time of cognitive restorati on (e.g., learning and memory consolidation). With increasing age, depth of sleep decreases both quantitativ ely and qualitatively (Morgan, 2000). Older adults sleep is structurally lighter because they spend less time in stages 3-4 of sleep and more time in stages 1-2. Riedel and Lichstein (1998) used a correlational, crosssectional design and two nights of polysomnogra phic data to examine the relationship between objective sleep measures and subjective sleep satis faction in a sample of 47 older adults with primary insomnia. Results indicated that depth of sleep (increased time spent in stages 3-4 and decreased time devoted to stages 1-2) and sleep la tency (time required to initiate sleep) were the strongest predictors of subjective sleep satisf action. Therefore, structural sleep changes associated with normative aging influence sleep in a manner that increases nonrestorative sleep and decreases sleep quality (Riedel & Lichstein, 1998). 23

PAGE 24

Second, duration of sleep decreases gradua lly, throughout the life span, from infancy to late life. Secondary to decrea sed total sleep time, decreased total REM time and decreased sleep efficiency also occur (Morgan, 2000). Often, tim e spent in Stage 4 decr eases to the point of disappearing entirely. Third, contin uity of sleep also deteriorates with age, as older adults shift among sleep stages more frequently than younger individuals such that somatic and cognitive restoration occurs in increasin gly smaller segments throughout the night. Age also brings increased wakefulness during the sleep period and intrasleep arousals (Morgan, 2000). Since increases in wakefulness and increases in shifts among stages make sleep unconsolidated sleep, it follows from this that less time is spent in slow wave sleep and sl eep satisfaction may be compromised (Morgan, 2000). Psychological Distress, Cardiac Disease, and Sleep Decades of research have found a strong relationship between cardiac disease and psychological distress. People with CAD are more likely to suffer from depression than otherwise healthy people; similarly, people with depression are at greater risk for developing heart disease (NIMH, 2002). One in five pa tients with newly diagnosed CAD has major depression, and one in three individuals who ha ve survived a heart attack are depressed (Glassman & Shapiro, 1998). This is problematic in light of the fact that depression may potentiate the effects of other ri sk factors for cardiac events (C arney, Freedland, Miller, & Jaffe, 2002) and has been associated with a 59% increase in mortality risk during a one-year follow-up (Rovner et al., 1991). When Cross and colleagues (2006) examined a heterogeneous sample of 69 cardiac patients, clinically significant depressive symptoms per the Center for Epidemiological Studies Depression Scale (CES -D) were significantly associated with poorer sleep patterns. Across a one-week period, depressed patients had shorter mean total sleep time (6h, 31m + 100m versus 7h, 22m + 56 m; F (1,67) = 2.16; p = 0.035) and had a longer mean 24

PAGE 25

sleep latency (28m + 28m versus 18m + 12m; F (1,67) = 2.14; p=0.037) than their nondepressed counterparts. These results suggest that depr ession and cardiac disease may work in tandem to correlate with poorer sleep patterns. Depression is also a common concern in ICD patients. Konstam (1995) measured depression in ICD patients and found that nearly a quarter (24%) was ex periencing clinically significant depression. Hegel and colleagues (1998) examined patients across several years of ICD therapy and found that 33% reported clinic ally significant depression, and the authors concluded that there is little reason to belie ve that depressive symptoms resolve without treatment as time since implantation increases. As part of the Tri ggers of Ventricular Arrhythmias (TOVA) study, Whang and colleagues (2005) examined depression in ICD patients and found it to be a significant and independent predictor of time until shock discharge since study entry. These results held when covarying fo r several medical and lifestyle factors such as ejection fraction, underlying cardiac disease, number of previous shocks, BMI, and tobacco use. Taken together, these results indicate that 24 33% of ICD patients experience depressive symptoms at clinically significant levels (S ears et al., 1999; Sears & Conti, 2006). Although these rates are not higher than those found in other cardiac patients, depr essed ICD patients are more likely to be shocked (Whang et al., 2005). Behavioral mechanisms for depression in cardiac patients include loss of physical functioni ng, changes in familial and social roles, and sexual dysfunction. Physiological mechanisms include altered cardiac autonomic tone, platelet reactivity, chronic inflammation to the vascular endothelium, and nonadherence to prescribed cardiac regimens (Carney, Freed land, Miller, & Jaffe, 2002). Anxiety is also frequently comorbid with cardiac disease. In the general population, generalized anxiety disorder is associated with an elevated risk for CAD (Barger & Sydeman, 25

PAGE 26

2005). The Framingham Offspring study (Eak er, Sullivan, Kelly-Hayes, DAgostino, & Benjamin, 2005) found that anxiety is a risk factor for total mortality in men ( RR = 1.22; 95% CI, 1.08 1.38) and women ( RR = 1.27; 95% CI, 1.05 1.55). Although anxiety is often found to be associated with coronary disease, evidence to the contrary also exists. For example, Suls and Bunde (2005) found in their qua litative review of 17 published re ports that only four studies found positive associations between anxiety at baseline and subsequent cardiac morbidity or mortality. This finding suggests a limited link between anxiety and coronary disease. Anxiety tends to be more of a salient factor in patients with cardi ac arrhythmias who have ICDs. Herrmann et al. (1997) measured anxiety in ICD patients using the Hospital Anxiety and Depression Scale (HADS) and found that 13% of patie nts reported clin ically significant levels of anxiety. Schuster et al. (1998) used the State-Trait Anxiety Inventory and found that 38% of patients had clinically significan t levels of anxiety. Pycha and colleagues (1990) found that hypervigilance is common in ICD patients, and 94% of patients endorsed a preoccupation with their cardiac functioning in th e years following implantation. In summary, 13-38% of ICD patients exhibit clinically significant anxiety, an d these symptoms frequently relate to shock anxiety and fears of device malfuncti on and death (Sears et al., 1999). Unfortunately, insomnia is also comorbid with both anxiety and depression. In patients with insomnia, 14-20% show evidence of majo r depressive disorder (Mellinger, Balter, & Uhlenhuth, 1985; Ford & Kamerow, 1989). Lifetime prevalence estimates of sleep disturbance and psychiatric disorders reveal that 31.1% of those with inso mnia have experienced major depressive disorder (Breslau et al., 1996). In patients presenting to general medical clinics, sleep disturbance had a large positive predictive valu e (61%) for clinically significant depression (Gerber, Barrett, & Barrett, 1992). Key polys omnographic findings that associate poor sleep 26

PAGE 27

with depression have also been found. For instan ce, depressed individuals show decreased sleep continuity via increased sleep latency, increa sed wakefulness during sleep, and early morning awakenings (Benca, 2005). Decreas es in slow wave sleep have been found in depressed patients (Kupfer, Reynolds, & Ulrich, 1986). Lastly, REM abnormalities have been reported in the form of decreased REM sleep latency, an increased per centage of REM sleep, and an increased rate of rapid eye movements (Benca, 2005). Patients may report these REM abnormalities as having disturbing dreams (Benca, 2005). These change s in sleep architecture may relate to the underlying neurobiology of depr ession (Benca, 2005). Much like depression, anxiety and insomnia commonly co-occur and manifest observable changes to sleep architecture. Compared to h ealthy controls, patients with Generalized Anxiety Disorder have increased sleep latency, increase d waking after sleep onset decreased total sleep time, and decreased sleep efficiency (Monti & Monti, 2000). Various forms of anxiety, including post-traumatic stress disorder, panic disorder, and social phobia, manifest sleep problems through specific mechanisms but with the common thread of somatic and/or cognitive hyperarousal that is incompatible w ith sleep (Stein & Mellman, 2005). Insomnia is frequently comorbid with anxi ety and depression, and this is especially pertinent in the assessment and treatment of wo men with cardiac disease. Women experience depression twice as often as men, and the lifetime prevalence for depression is 20% in women compared to 10% in men (Bhatia & Bhatia, 1999 ). Anxiety is also more prevalent among women. For example, women have a two to th ree times greater lifetime risk than men of developing generalized anxiety disord er, and their lifetime prevalence of developing post-traumatic stress disorder (P TSD) is twice that of men (P igott, 2002). In women with coronary artery disease, 28-42% report sympto ms of insomnia (e.g., difficulty initiating and 27

PAGE 28

maintaining sleep) whereas only 8-33% of male s report these symptoms (Edell-Gustafssoon, Svanborg, Swahn, 2006). Insomnia is associated with poor quality of life, and it is also a risk factor for a variety of othe r physical and psychological condi tions (Taylor, Lichstein, & Durrence, 2003). In summary, anxiety and depr ession are both significantly comorbid with cardiac disease and insomnia. Depression tends to be associated more with coronary disease, while anxiety is more closely related to cardiac arrhythmias and ICD therapy. Depression, anxiety, and sleep patterns can be bidirectionally detrimental among one another, and depression is especially prevalent in insomnia patients. Female gender may pose additional risks in terms of psychological distress. Actigraphy Studies on Sleep in Chronic Disease Sleep is a multidimensional construct that can be measured in terms of consolidation, duration, and quality. Many forms of sleep disord ers require referral to a physician specializing in sleep medicine for assessment and pharmacological treatment. The following is a description of the ways in which sleep is measured through objective indices, subjective indices, and sleep diaries. Polysomnography (PSG) involves the all-night, continuous sleep recording of multiple variables during sleep. The gold standard of sleep measures, polysomnography is the most accurate quantitative measure of sleep and it allows for the analysis of physiological activity during specific sleep stages. Despite the wealth of information that a polysomnogram provides, this method analyzes a very limited sample of sleep and is often uncomfortable for patients. Further, polysomnography is not practical for po pulation studies with disease groups due to the high financial cost. 28

PAGE 29

The American Sleep Disorders Associati on (1997) indications for polsomnography evaluation include: (1) suspected obstructive sleep apnea (excessive daytime sleepiness, snoring, or pauses in breathing) or CPAP titration, (2 ) periodic limb movements, (3) narcolepsy (combined with a Multiple Sleep Latency Test), (4) sleep-related behaviors that are violent or potentially injurious to the patients or others. Polysomnography is not indicated for insomnia for several reasons (ASDA, 1997). These reasons include the notion that sleep can improve or worsen due to sleeping in an unf amiliar environment and the fact that insomnia typically varies in severity across nights so that a single night may not accurately depict the full extent of sleep problems (AASM, 2003). Actigraphy is a method of measuri ng sleep that is based on the f act that during sleep there is little movement activity, whereas during wake there is increased movement (Ancoli-Israel, 2005). In other words, gross motor movement is used as a proxy for distinguishing between sleep and wakefulness. An actigraph is a lightwe ight, watch-like device that is usually worn on the nondominant hand. The device contains a mo vement detector (e.g., accelerometer) with ample memory for recording long periods of time. Movement is usually sampled several times per second and stored in 30-second epochs. Advantages of using actigraphy in clude observing sleep in the na tural setting, the ability to monitor sleep patterns over a prolonged period of time without needing to download data, and lower costs in comparison to polysomnography. C oncordance rates for distinguishing sleep from wake using actigraphy versus polysomnography have been high at 82 to 95% (Cole, Kripke, Gruen, Mullaney, & Gillin, 1992; Blood, Sack, Pe rcy, & Pen, 1997). Recent research on a sample of 57 adults with insomnia provide d evidence that validated actigraphy against polysomnography on 4 out of 5 sleep parameters: number of awakenings, wake time after sleep 29

PAGE 30

onset, total sleep time, and sleep efficiency (L ichstein et al., 2006). Actigraphy measures these sleep parameters more accurate ly than waking after sleep onse t, although the device provides useful information for all of thes e indices (Cole et al., 1992; Buysse et al., 2006; Lichstein et al., 2006). Actigraph data is highly reliable and valid for use with healthy populations, although its reliability decreases as sleep becomes more disturbed (Ancoli-Israel, 2005). Summary and Hypotheses Regarding Sleep and Cardiac Disease The prevalence and broad impact of cardiac disease have urged researchers to manage the full spectrum of behavior in disease manageme nt. Although the importance of understanding the influence of sleep patterns on car diac disease has been highlight ed by recent research, little information is known regarding how objective and subjective measures of sleep patterns compare to one another and which psychological variables are most strongly related to poor sleep in cardiac samples. The proposed study wi ll shed light on the feasibility of using subjective indices to measure sl eep in the cardiac population based on the degree of consistency between subjective and objective sleep data in two subgroups of patients. This study will also help us to understand how sleep patterns may differ between patients with classic forms of cardiac disease (CAD) versus electrophysiology (ICD) patients. Additionally, the proposed research can help us to understand the degree to which physi cal limitations posed by cardiac conditions can be associated with maladaptive sleep. 30

PAGE 31

Aim #1 : Do ICD patients differ from CAD patients across a 14-day period on standard sleep indices (sleep efficiency, sleep onset latency, total sleep time, sleep quality, and waking after sleep onset) per sleep diary data and per actigraphy data? Hypothesis #1: ICD patients will have sleep indi ces consistent with poorer sleep per sleep diary data and per actigraphy data related to hypervigilance for device functioning and nocturnal shock discharge. Aim #2 : What are the predictors of sleep e fficiency in the overall cardiac sample? Hypothesis #2: Age, sex, left ventricular ejec tion fraction, psychological distress, somatic hypervigilance, daytime physical act ivity levels, and other standard sleep indices will strongly predict sleep ef ficiency in the overall study sample. Table 1-1. Commonly Used Sl eep Indices (Specific to Sl eep Diary Data Collection) Sleep Onset Latency (SOL) Period of time in minutes that elapses between turning off the lights with the intention of initiating sleep and the beginning of the first stage of sleep Sleep Efficiency (SE) Comparison of the number of hours in bed that are spent actually sleeping with the total time spent in bed, expressed as a percentage. 2 A sleep efficiency of 85% or higher is considered healthy or adaptive if it occurs in the absence of insomnia complaints and impaired daytime functioning Sleep Quality (SQ) Subjective statement regarding the degree to which sleep is restorative or non-disrupted Total Sleep Time (TST) Total time spent sleeping during a given night Waking After Sleep Onset (WASO) Time spent awake after init iating sleep until the last awakening 2 Sleep Efficiency is a percentage that is calculated by dividing the number of hours spent sleeping by the number of hours spent in bed, multiplied by 100. For example, a pa tient who spends 9 hours in bed but who only actually sleeps 6.5 of those hours would have a sleep efficiency of 72%. 31

PAGE 32

CHAPTER 2 METHODS Participants and Recruitment Two groups of participants were recruited: CAD patient and ICD recipients. Participants were recruited during their outpat ient consultation visit at the University of Florida & Shands Hospital Cardiovascular Medicine Clinics. Key inclusion criteria included the following: (a) diagnosis of coronary artery di sease or a cardiac diagnosis indi cating a need for ICD therapy, (b) 18 years of age and older, and (c) written and verb al fluency in English. Key exclusion criteria consisted of: (a) diagnosis of an intrinsic or extrinsic primary dyssomnia other than insomnia (including but not limited to OSA 3 central sleep apnea, restle ss legs syndrome, periodic limb movement disorder, narcolepsy, a nd circadian rhythm sleep disorder s) (b) diagnosis of a primary parasomnia (e.g., sleep walking, sl eep terrors), (c) current psyc hotic disorder or dementia disorder, and (d) current suicidal ideation or pl an. Patients who met st udy criteria and were taking sleep medications such as benzodiazepines we re included during the recruitment process. Recruitment for the proposed study was fac ilitated by the integration of the Cardiac Psychology Research Lab with the Electrophysiology (EP) service. The EP clinics of Jamie B. Conti, M.D., Thomas Burkart, M.D., and Sherry Saxonhouse, M.D. were attended in order to recruit ICD patients. At these clinics, both new and follow-up EP patients with ICDs are seen. The EP service at UF & Shands follows appr oximately 700 ICD patients per year, such that recruitment for the desired sample size was not problematic. Of the overall group of patients 3 Per criteria defined by the International Classification of Sleep Disorders Revised (ICSD-R) Diagnostic and Coding Manual, obstructive sleep apnea is defined as ha ving >5 obstructive apneas, each lasting >10 seconds, per hour of sleep, in conjunction with other accompanying symptoms. 32

PAGE 33

seen at these EP clinics, approximately 40% have ICDs, 30% have non-malignant arrhythmias and are not undergoing ICD therapy, and the rema ining 30% have only signs and/or symptoms related to cardiac arrhythmias (e.g., syncope premature ventricular complexes). The Interventional Cardiology Clinics of Scott Denardo, M. D. and Karen Smith, M.D. were attended in order to recru it CAD patients. Of the overall group of patients seen at these clinics, approximately 60% have corona ry artery disease without ICDs. The study was reviewed and approved by th e Gainesville Health Science Center Institutional Review Board-01 at the University of Florida. Participation was contingent upon receiving informed consent. All participants were treated in accordance with the Ethical Principles of Psychologists and Code of Conduct (American Psychological Association, 2002). Measures Demographic Information A Background Information Questionnaire was included in the pack et of psychological questionnaires. Information was gathered on gend er, age, educational level, occupation, marital status, religion, number of ch ildren, and financial income. Biomedical Data For each participant, information on cardiac diagnoses, EF%, and current medications was gathered via medical record review. EF% wa s used as a biomedical marker of the hearts pumping function for patients with ICDs. Specifically, this is the percentage of blood that the left ventricle ejects into the arteries per heartb eat. The mean EF% for the general population is 68%, with a standard deviation of 9% (Bra unwald, 1992). As values decrease, pumping inefficiency increases, and an EF% of < 35% is one criterion for ICD implantation (Bardy, 2005). Additionally, data on time since implantation an d the frequency of ICD shock discharge was gathered through inquiry of ICD patients. This information was compiled in order to measure 33

PAGE 34

the level of hyperarousal in these patients that may be influenced by shock. A two-year period of shock history was used as a proxy to total lifetime histor y of shock. For participants using sleep medications such as benzodiazepines, time sin ce initiating sleep medication use was also collected. Objective Sleep Data Actigraphy. The current study employed the Actiwatch-64 by Mini-Mitter/Respironics. This type of device records gross motor movement s in measuring sleep/wake cycles. This model of actigraphy has been used previously in st udies on insomnia, restless legs syndrome, and a variety of treatment outcome studies (Redek er, Mason, Wykpisz, & Glica, 1995; Redeker & Wykpisz, 1999; Ancoli-I srael et al., 2003). As described by Ancoli-Israel and colleagues (2003), actigraphy data can be digitized using three primary strategies: (1) time above threshold, (2) zero-crossing method, and (3) digital integration. Each of thes e strategies have their respecti ve strengths and limitations, and newer devices that use more than one method of acquiring data can help overcome these limitations. The actigraphs used in the current study utilized the digital integration method. This technique samples activity at a very high rate (such as 20 or more times per second). The average activity value recorded at each sample is stored and, in a given epoch, these values are used to derive the average activity level with in the epoch (Gomy & Allen, 1999). The area under the curve is calculated for each 30-second epoc h (Ancoli-Israel, 2005). Gorny and colleagues (2001) found that this method is su perior to the former methods of data collection for identifying movement amplitude. Actigraph devices can continuous ly record movement behavior for 24-hours daily for days, weeks, or even longer (Ancoli-Israel et al., 2003). 34

PAGE 35

Subjective Sleep Data Sleep diary data. Sleep diaries are a form of self-m onitoring for nightly sleep patterns. Participants completed a prospective, two-week Sleep Diary adapted from C. McCraes sleep diary containing standard sleep i ndices of bedtime, rise time, sl eep onset latency (SOL), number of awakenings (NWAK), wake time after sleep on set (WASO), and a sleep quality rating. Time in bed (TIB), total sleep time (TST), and sleep efficiency (SE) will be derived from this information. Data on naps, caffeine use, and bed time medication use will al so be gathered using this instrument. Sleep diary monitoring is the most standard, practical, use r-friendly, and cost-efficient method for repeated, accurate samp ling of target sleep behaviors (Lacks & Morin, 1992; Spielman, Yang, & Glovinsk y, 2005). Although data from sleep diaries do not always reflect absolute data from objective sleep measurements such as polysomnography and actigraphy, subjective estima tes of sleep patterns derived from sleep diaries provide a reliable and valid index of clinically significant sleep problems (Coates et al., 1982). Rogers, Caruso, and Aldrich (1993) found that the percentage agreement between th e subjective data recorded in the sleep diaries and polysomnographic data was acceptable (kappa = .81) in a sample of 50 sleep disorder patients and matched controls. Means and colleagues ( 2003) found that persons with insomnia have a tendency to both over and under-estimate their sleep time on sleep diaries. Fortunately, this tendency to overor under-estimat e is usually consistent within an individual (Means, Edinger, Glenn, & Fins, 2003). Sleep quality. The Pittsburgh Sleep Quality Index (PSQI; Buysse et al., 1989) was designed to measure sleep quality in clinical populations over a one-month period. The PSQI provides one global and seven component scores of sleep quality: subjective sleep quality, sleep latency, sleep duration, habitual sleep efficienc y, sleep disturbances, us e of sleeping medication, and daytime dysfunction. Global scores range from 0-21, with higher scores indicating poorer 35

PAGE 36

sleep quality. In disti nguishing good sleepers from poor sleepers a total score of 5 resulted in a diagnostic sensitivity of 89.6% and a specificity of 86.5% ( kappa = 0.75, p <0.001). This instrument has also demonstrated good internal consistency reliability ( Cronbachs alpha =.83; Carpenter & Andrykowski, 1988). Serber and coll eagues (2003) used the PSQI to examine sleep quality in ICD patients and found that the mean PSQI total score was 7.99 ( SD = 2.60), with 87% of the total population scoring above 5 for global sleep quality. In the current study, the PSQI global score was used as a measure of sleep qua lity. Additionally item #8b, an item that does not contribute to the total score, was used as a measure of daytime dysfunction. Mood and anxiety. The Hospital Anxiety and De pression Scale (HADS; Zigmond & Snaith, 1983) was designed to identify probable and possible caseness of anxiety disorders and depression in medical patients. Somatic symptoms of anxiety and depressive disorders, such as insomnia, dizziness, headaches, and fatigue are not included such that the items distinguish anxiety and depression from physical conditions. Although th e HADS total score is most commonly used, there are also individual scales for the two constructs of anxiety (HADS-A) and depression (HADS-D). Bjelland an d colleagues (2002) reviewed 747 studies that used the HADS and found that it demonstrated good internal consistency reliability ( Cronbachs alpha for HADS-A = .68-.93; for HADS-D = .67.90), good convergent validity with the Beck Depression Inventory-II and the State Trait Anxiety Inventory ( r = .61-.83), and excellent sensitivity (.83.90) and specificity (.78-.79). There is evidence for both a 2-factor structure of anxiety and depression and a 3-factor struct ure of autonomic anxiety, negativ e affect, and anhedonic affect; both of these models have demonstrated a similar goodness-of-fit (Dunbar, 2000; Lisspers, 1997; Spinhoven et al., 1997). The curre nt study used the HADS total scor e as a composite measure of anxiety and depression, although subscale score for anxiety and depression will also be 36

PAGE 37

examined. Consistent with Snaith (2003), a score > 11 will be considered clinically significant for both the anxiety a nd depression subscales. Somatic hypervigilance. The Body Vigilance Scale (BVS; Schmidt, Lerew, & Trakowski, 1997) was designed to measure attentional focus on interoceptive activity. This 4-item selfreport inventory measures vigilan ce for somatic symptoms on a 10-poi nt scale. The last item in this scale gives ratings for attention given to se nsations during a variety of symptoms, including heart palpitations, chest pain/discomfort, shortness of breath, and faintness. The BVS has good internal consistency and test-r etest reliability. It is psychometrically sound for use with both clinical and nonclinical populations, and a pr incipal components analysis revealed that this construct is unifactorial in both settings. Higher scores on the BVS are associated with anxiety sensitivity, and anxiet y symptoms, including spon taneous panic attacks (Schmidt, Lerew, & Trakowski, 1997). In Schmid t, Lerew, & Trakowskis (1997) sample of 71 community residents, the total BVS score was 18.3. For panic disorder patients ( n = 48), the mean total BVS score was 22.6 and for social phobia patients ( n = 18) it was 17.6. Functional status in cardiac disease. The Duke Activity Status Index (DASI; Hlatky et al., 1989) assesses self-reported functional capacity and some aspect s of quality of life. The 12item instrument measures major activities of daily living, including personal care, household tasks, recreational activities, pl aying sports, and sexual functi oning. Each item is weighted differently based on the known metabolic cost of each activity, and weights of positive terms are summed to form the total DASI score (Koch et al., 2004). Scores range from 0 to 58.2, with higher scores indicating healthie r physical functioning. Total DAS I scores correlated well with peak oxygen uptake in two independent samples of N = 50 each ( Spearmans rho = .58 and .91; p< 0.001). 37

PAGE 38

Shock anxiety. The Florida Shock Anxiety Scale (FSA S; Kuhl, Dixit, Walker, Conti, & Sears, 2006) was designed to assess common worri es and anxiety related to experiencing ICD shock discharge. It measures the cognitive, behavioral, emotional, and social impact of shock anxiety in 10 items. The FSAS has good reliability (Cronbachs alpha = 0.91, split-half = 0.92), and it has demonstrated good convergent validity with the Multidimensional Fear of Death Scale ( r = -0.65). Factor analysis rev ealed a 2-factor structure consis ting of a consequence factor (e.g., fear of creating a scene) and a tr iggers factor (e.g, getting angr y or engaging in sexual activity; Kuhl et al., 2006). The current study used the FSAS total score in measuring shock anxiety. Angina. The Seattle Questionnaire (SAQ; Spertus et al., 1995) was designed to measure five clinically important dimensions of health in patients with coronary artery disease. These dimensions relate to the constructs of angina chest pain, and chest discomfort. They are physical limitation, anginal stability, anginal frequency, treatment satisfaction, and disease perception. Higher scores indicat e less difficulty with chest pai n, and there is no one summary score. The validity of the SAQ has been establis hed by correlating its scales with various other measures of diagnosis and CAD patient functi oning including physician diagnoses, number of nitroglycerin refills, the Duke Activity Status Index, the SF-36, the exercise treadmill test, the Canadian Cardiovascular Societ y Classification, the Specific Activ ity Scale, and the American Board of Internal Medicine Patient Satisfaction Questionnaire. The five SAQ scales correlate significantly with these other measures, r = 0.31 0.70, p < 0.001. Procedures The author attended Cardiology clinics to id entify potentially elig ible participants through consultation with the attending physicians, fellows, and nursing staff. A member of the clinic staff introduced the study to the patient. If the patient was amenable, a research staff member came to the patients consultation room in order to give him or her more information 38

PAGE 39

about the study. If the patient expressed interest in study participation, he or she reviewed the Institutional Review Board-approv ed informed consent form and was given an opportunity to ask questions concerning study procedures. Each pa rticipant was informed that study participation was both voluntary and confidential. Next, a brief screening inte rview was conducted in order to confirm that the patient met inclusion criteria an d did not meet any of the exclusion criteria. A plan was in place to refer patients experienci ng suicidal ideation/plan to the UF & Shands Psychiatry Service via the Emergency Departme nt immediately, although there was never need to implement this. Once informed consent was obtained (on Day 1), participants were assinged to either the Sleep Diary Only arm or the Sleep Diary + Actigraphy arm of the study. Assignment to group was based on resource accessibility (e.g., availabi lity of actigraphs when needed). This method was necessary to assure the flow of incoming data as some participants were unable to return their actiwatches as soon as planned upon their study completion a nd as two participants never returned their actiwatches to the author. Participants completed their packet of psychological self-repor t questionnaires and returned the data on Day 1. Although participants were encourag ed to remain in clinic long enough to complete the psychological questionnaire s, a subset of partic ipants was unable to complete the questionnaires in clinic due to time constraints. Th ese participants were instructed to complete the questionnaires at home on Da y 1 and return the measures by mail along with their sleep diaries. Participants completed thei r sleep diaries for Days 1-14. Sleep diaries were returned in a stamped, self-addressed envelope pr ovided to the participant. Study participation ended upon receipt of both the psychologi cal questionnaires and sleep diaries. 39

PAGE 40

All participants were asked to complete que stionnaires on sleep patterns (diary data), sleep quality, depression and a nxiety, somatic hypervigilance, and daytime physical activity levels. In addition, ICD patients were asked to complete a measure on an xiety related to shock discharge and CAD patients were asked to complete a measure on chest pain. The participant burden to complete these psychological questionnaires was low, as most patients took 15-30 minutes to complete them. Sleep diaries usua lly take patients approxi mately three to five minutes per day, for 14 days, to complete. Participants received two phone calls fr om the author during Days 1-14 of data collection. The first phone call took place during th e first few days of the first week of data collection; the second phone call took place duri ng the second week of data collection. The purpose of these calls was twofold. First, callin g participants encouraged compliance with both sleep diary completion and actigra phy data collection. Second, these phone calls provided an opportunity for participants to voice questions or concerns regard ing data collection. A subset of participants ( n = 14 CAD patients and n = 15 ICD patients) used actigraphy in order to obtain an objective index of their sleep patterns. Partic ipants were instructed to wear their actigraph device th roughout the 24-hour period for Days 1-14, with the exception of during showering and swimming. The actigraphs were c onfigured to begin data collection at 17:00 on Day 1 of data collection. Sleep diaries were gene rally used as a validity check for actigraph data. Since the battery life for the actigraphs was approximately 30 days, all actigraphy data was retrieved at once upon a participan t returning the actigraph. Partic ipants were instructed to return their devices at the e nd of the 14 days via mail, by br inging them the Cardiology West outpatient clinic, or by bringing them by the UF & Shands Psychology Clinic. At study 40

PAGE 41

completion, all participants were given a short booklet on healthy sleep habits and patterns compiled by the author. Sleep diary data was used to anchor bedtimes and rise times in the actigraphy data. The following time intervals in the actigraphy data were excluded from analyses. First, time periods in which it appeared that a pa rticipant removed his or her actiw atch for longer than 60 minutes were excluded from analyses. These time periods were blocks of complete inactivity that did not correspond to times for watch removal (during s howering or swimming) or naptimes. Second, Day #14 of actigraphy data was excluded from anal yses, as there was no sleep diary data to accompany this portion of the actogram. This resu lted from using a sleep diary that probed for 14 days worth of data collection but that used y esterday as the first day of recalling sleep data for recording Statistical Analyses Data were analyzed using Actiware versi on 5.0 (Mini-Mitter/Respironics, Bend, OR) and SPSS version 15.0 (SPSS, Cary, NC). Standard mean sleep indices of sleep efficiency, sleep onset latency, total sleep time, and waking after sl eep onset were calculated for each participant across the 14-day data collection period. Aim 1 of the current study involved examining group differences between CAD and ICD patients on thes e standard sleep indices using subjective and objective data collection methods across the 14-day study period. Independent Samples t -tests were used to assess the differences between CAD and ICD patients for both sleep diary-based sleep indices and actigraphybased sleep indices. Aim 2 involved examining predictors of sleep efficiency in both the CAD and ICD groups. Sleep efficiency was chosen as the crit erion variable for two reasons. First, sleep efficiency is more informative than some other sleep indices such as total sleep time, because there is increased interindividual variability and a smaller range of values representing adaptive 41

PAGE 42

sleep in sleep efficiency. Second, sl eep efficiency is a useful variable in that it is used to titrate sleep restriction, an important behavioral sl eep intervention (Perlis 2006). To examine predictors of sleep efficiency, a hierarchical regression analysis was employed. Psychological variables and physical activity data were used to predict the criteri on variable of sleep efficiency. Controlled medical variables includ ed age, sex, and left ventricula r ejection fraction. First, age, sex, and left ventricular ejection fraction was entered into the model. Second, psychological variables including distress ( HADS total anxiety and depres sion score) and somatic hypervigilance (BVS scores) were entered as a block into the m odel. Third, daytime physical activity (DASI scores) was entered into the model. Fourth, history of ICD shocks was entered into the analysis. This hierarchical model was hypothesized to allow the model to determine the relative salience of each type of pr edictor in forecasting the criteri on variables. One hierarchical regression was conducted for the overall study sample. Often, multicollinearity is a concern in regres sion analyses when predictor variables are used in measuring similar latent constructs. Multicollinearity occurs when individual predictor variables are highly correlated and share much of the same pool of variance in the overall process of explaining the variance in the criterion variable. Multicollinearity can be problematic in that slight fluctuations in th e data can lead to substantial fluctuations in the size and even the valence of statistical outcomes. The proposed st udy decreased the threat of multicollinearity and increased the proportion of unique variance by examining Pearson Product-Moment correlations among total sleep time, sleep onset latency, waking after sleep onset, and sleep quality. Of these variables, sleep indices that correlated at r = > 0.7 were excluded from the sleep indices block of the regression analyses used to predict sleep efficiency. 42

PAGE 43

Power and Sample Size Calculations GPOWER computer software (Faul & Erdfel der, 1992) was used to calculate a power analysis to determine the sample size necessary for accurate and reliable statistical judgments. In previous research conducted in this research lab using a sample of N = 100, the standard deviation for mean sleep efficiency was 12.37 for ICD patients and 10.62 for CAD patients. The standard deviation for mean total sleep time was 108 minutes for ICD patients and 87 minutes for CAD patients. Considering a minimum nontrivial effect size of 5% for sleep efficiency, Cohens d = 0.40 for ICD patients and d = 0.47 for CAD patients ( M of ds = 0.435). Considering a minimum nontrivial effect size of 30 minutes fo r total sleep time, Cohens d = 0.27 for ICD patients and d = 0.34 for CAD patients ( M of ds = 0.31). Averaging these two Cohens d values yielded a value of 0.37. Using this value of 0.37 in the es timation of statistical power for the proposed study, it calculates that power = 64% at p < 0.05 (one-sided) for a total sample size of 60. Table 2-1. Psychosocial Measures Predictor Variables: 1) Standard Sleep Variables ( Objective ) Actigraphy (Actiwatch-64 by MiniMitter/Respironics, seven (7) devices for use) 2) Standard Sleep Variables ( Subjective ) Sleep Diaries: version based on C. McCraes sleep diary containing standard sleep indices Sleep Quality: Pittsburgh Sleep Quality Index (PSQI; Buysse et al., 1989) 3) Functional Status: Physical Activity: Duke Activity Status Index (DASI; Hl atky et al., 1989) 4) Psychological Adjustment: Depression and Anxiety: Hospital Anxiety and Depression Scale (HADS; Zigmond & Snaith; 1983) Somatic Hypervigilance: Body Vigilance Scale (BVS; Schmidt et al., 1997) Shock Anxiety: Florida Shock Anxiety Scale (FSAS; Kuhl et al., 2006) Criterion Variable: 1) Sleep Efficiency: Per Actigraphy data Per Sleep Diary data 43

PAGE 44

CHAPTER 3 RESULTS Eighty-three patients at the Cardiology clin ics of UF & Shands met study criteria and were approached regarding st udy participation. Sixty-four of these patients agreed to study participation and 19 declined par ticipation. Four (6%) of these 64 patients were fully lost to follow-up. Four other participants (6%) provided onl y partial data sets: tw o participants did not provide any sleep diary data but completed psychosocial measur es and/or actigraphy data, one participant did not provide psychosocial data bu t did complete sleep diary and actigraphy data, and one participant provided sleep diary and psychosocial data but did not provide actigraphy data. Of this final sample of 60 participan ts, 30 (50.%) were ICD patients and 30 (50.%) were CAD patients; 31 (52%) comprised the Sleep Diary only arm and 29 (48%) comprised the Actigraphy plus Sleep Diary arm. Demographic and Medical Variable Descriptives and Cardiac Group Differences Means and standard deviations for characteri zing study participants are provided in Table 3-1. Most patients were male, Caucasian, married, Protestant, and either working full-time or on disability. There were no group differences found between the ICD and CAD groups for any of the demographic variables including age, sex, race marital status, number of children, religion, employment status, or income. ICD and CAD specific descriptive data is pr esented in Table 3-2. Most of the ICD patients (76.7%) had a single or dual-chambered device; 23.3% of individuals had biventricular devices. Among ICD patients, over half (58.1%) had underlying coronary disease, with 29.0% having a history of myocardial infarction and 38. 7% having a history of coronary artery bypass surgery. Among CAD patients, nearly half had a hi story of myocardial in farction and nearly as many (41.4%) had a history of coronary artery b ypass surgery. Of the 30 defibrillator patients, 44

PAGE 45

nine (30%) had experienced ICD shock discharge in the past two years. Of these nine shocked patients, all but one received 17 shocks in the previous two y ears; one participant received 25 shocks during this time period indi cating inappropriate shock delivery. When the ICD and CAD groups were compared for medical variables, no group differences were found for the lik elihood of having orthopnea, hist ory of myocardial infarction, or history of coronary artery bypass graft surg ery. Similarly, no group differences were found for nearly all medication use (including hypnotic use, time since begi nning hypnotic medication, beta-blocker use, gastroesophageal reflux disease medication use, antidepressant use, anxiolytic use, or use of other medications that are known to affect sleep) (Buy sse, Schweitzer, & Moul, 2005; Nishino & Mignot, 2005; Schweitzer, 2005). However, ICD patients (61.3%) were more likely than CAD patients (31.0%) to be taking diuretic medication, phi coefficient = .019. As expected, the mean ejection fract ion % for ICD patients (35.53%) was lower that that of CAD patients (48.69%). Psychosocial and Sleep Variable Descri ptives and Cardiac Group Differences Overall, the sample exhibited a low level of anxiety and depression per HADS scores (M = 10.68, SD = 7.99) and a high level of soma tic hypervigilance on BVS scores ( M = 21.17, SD = 9.41). They reported a high level of physical activity (per DASI scores M = 25.44; SD = 12.81) in relation to initial validation studies for this measure. Shock Anxiety scores in ICD patients were similar to those found in the initial validation study for this measure ( M = 17.31, SD = 8.56). ICD and CAD patients were also compared on psychosocial variables. Analyses of covariance using EF% as the control variable were performed for DASI and PSQI scores, as these distributions met the appropriate assu mptions of normality, homogeneity of variance, linearity, and homogeneity of re gression slopes. There were no differences for any of these 45

PAGE 46

measures. For HADS scores, Shapiro-Wilks W tests revealed nonnormality in kurtosis. For BVS scores, Levenes test revealed a lack of homogeneity of varian ces. Therefore, Kruskal-Wallis ANOVAs were conducted for these three variables. Similarly, no cardiac group differences were found for any of these measures. Regarding HADS scores, two partic ipants had outlying scores of z > 3.00. Therefore, these data were excluded from analyses. HADS da ta were also analyzed for levels of anxiety and depression that would be considered clinica lly diagnostic, defined as a subscale score >11 for both the anxiety subscale and the depres sion subscale. While many patients reported symptoms of anxiety and depression suggesting mild levels of distress, only one ICD patient and three CAD patients reported moderate to severe levels of anxiety. Li kewise, only one ICD and two CAD patients reported moderate to severe le vels of depression according to interpretive guidelines for the HADS. Given the well-establis hed high prevalence of depression and anxiety found in cardiac populations, these fi ndings are surprising and incons istent with what would be expected. Analyses were also performed in order to examine differences between actigraphy and sleep diary indices (for the overa ll sample of combined ICD and CAD patients). See Table 3-3. For the sample of 29 participants from th e Actigraphy plus sleep diary arm, paired t -tests were used to compare means for sleep efficiency and to tal sleep time. Participants had significantly higher sleep efficiencies according to their sleep diaries (80.65%) compared to actigraphy (76.03%), t (58) = 2.43, p = .022. Participants also had si gnificantly higher total sleep times according to their sleep diaries (6.84 hours; 410.69 minutes) compared to actigraphy (6.40 hours; 384.29 minutes), t (58) = -25.29, p < .001. These results are contra ry to what would be expected 46

PAGE 47

given the tendency for actigraphy to overscore total sleep time which in effect generally increases sleep efficiency (Lichsetein et al., 2006). Waking after sleep onset (sleep diary) and sleep onset latency (sleep diary) did not meet normality criteria for kurtosis necessary for using t -tests, therefore a nonp arametric test was used. For waking after sleep onset, actigraphy estim ates were higher than sleep diary estimates, although the Sign test revealed that this difference did not reach statistic significance, (actigraphy M(sign) = 55.86, p = .17; sleep diary M(sign) = 30.54, p = .17). There were no significant differences for sleep onse t latency on actigraphy ( M (sign) = 30.14, p = .12) and sleep diary measures ( M (sign) = 32.66, p = .12). Number of awakenings after sleep onset was not compared between the objective and subjectiv e sleep measures, as this variable is computed in two different ways and has varying meaning be tween sleep diary and actigraphy measures. Aim 1: Assessment of ICD a nd CAD Group Differences for Sleep Diary and Actigraphy Data The first aim of the study was to determine if group differences exist between CAD and ICD patients for the two types of sleep indices, sleep diary data and actigraphy data. This was accomplished using a series of one-way analyses of covariance. EF% was used as a covariate, as this measure differed significantly between cardiac groups. Prior to performing these analyses of covari ance, data were tested for the assumptions necessary to perform these tests in order to refrai n from biasing parametric tests. Data collection involved randomness and independences of obser vations, and data were subjected to tests assessing normality and homogeneity of variance. Data were also analyzed for linearity. While formal tests for linearity exist, there is little agreement on which tests to use and how to interpret them (Tabachnich & Fidell, 2001). Therefore, linearity was examined by visual analysis of scatterplots on each of the dependent variables and by detecting outliers. One participant was 47

PAGE 48

found to be an outlier in that th e individual had scores of 3.00 > z > +3.00 from the mean for both waking after sleep onset (sleep diary) and sleep efficiency (s leep diary). Therefore, this data was excluded from analyses. Data were also analyzed for homogeneity of regression slopes by measuring the interaction term between cardi ac condition and EF%. Four va riables failed to meet the assumption of homogeneity of regr ession slopes and therefore analys es of variance were used for these dependent measures [for mean sleep effi ciency (actigraphy), mean total sleep time (actigraphy), mean waking after sleep onset (actigraphy), and mean total sleep time (sleep diary)]. For these four variables, Levenes technique was used to test the assumption of homogeneity of variance, and all da ta satisfied this assumption. As waking after sleep onset (sleep diary) and sleep onset latency (sleep diary) failed to meet the assumption of normality in kurtosis, nonparametric Kruskal-Wallis one-way ANOVAs were conducted for these dependent measures. In orde r to address an inflated Type I error risk in conducting this series of anal yses on group differences, a Bonf erroni-corrected level of p < .025 was employed for each pair of sleep indices (e.g., sleep efficiency for sleep diaries and actigraphy; total sleep time for sleep diaries a nd actigraphy). ANCOVA results are presented in Tables 312 through 3-16. There was no significant difference in how efficiently CAD patients sl ept (83.95%) compared to ICD patients (82.26%) according to sleep diary data, ( F (1,27) = 4.226, p = .046). Results did not reveal significant differences between the groups for any other variables e ither, including mean sleep onset latency (actigraphy), mean number of awakenings (actigraphy), mean sl eep efficiency (sleep diary), mean number of awakenings (sleep diary), and mean sleep quality (sleep diary). 48

PAGE 49

ANOVA results are presented in Tables 3-17 through 3-20. Significant differences were found for one actigraphy measure. CAD patients had much lower mean sleep efficiencies at 69.76% than ICD patients (82.80%), F(1,27) = 16.840, p < .001. This finding is surprising given that CAD patients had higher EF%s. This finding cannot be accounted for by differences in psychological distress, somatic hypervigil ance, or physical activity levels. Significant differences were al so found for one sleep diary m easure. CAD patients also had shorter mean total sleep times per sleep diaries (336.19 minutes; 5.60 hours) compared to ICD patients (430.65 minutes; 7.18 hours), F( 1,27) = 15.908, p < .001. No significant differences were found for mean waking after sleep onset (actigraphy) mean number of awakenings (sleep diary), or mean total sleep time (sleep diary). An analysis of variance using the Kruskal-Wa llis test (see Tables 3-21 and 3-22) showed that sleep onset latency (sleep diary) did not signifi cantly differ between ICD and CAD patients ( H 1,57 = 0.79, p = 0.38). Similarly, no si gnificant difference was f ound for waking after sleep onset (sleep diary), H 1,56 = 0.76, p = 0.38. Aim 2: Prediction Model Correlations among the full set of sleep a nd psychosocial variables can be found in Tables 3-4 and 3-5. Next, correlations are pres ented for the full sample of actigraphy data and sleep diary data separately (see Tables 3-6 and 3-7), illustrating relati ons among predictors and the criterion variable of sleep efficiency. Thir d, correlations of predicto rs and sleep efficiency are presented for CAD and ICD sample separately (Tables 3-8 through 3-11). Of note, Tables 3-6 through 3-11 exclude predictors from Tables 3-4 and 3-5 that correlated r > 0.7 with sleep efficiency. As expected, a wide range of significant correlations were found between and among psychosocial variables and sleep variables. 49

PAGE 50

Hierarchical Regression Models The second aim of the study is to examine predictors of sleep efficiency in the overall cardiac sample. An understanding of variables that predict sleep efficiency is important because this index is not biased by interindividual differe nces in sleep needs and is commonly used as an outcome measure in behavioral sleep treatmen ts. It was hypothesized that age, sex, left ventricular ejection fraction, ps ychological distress, somatic hypervigilance, daytime physical activity levels, and corresponding slee p indices would significantly pr edict sleep efficiency in the overall study sample. Age and sex were ultimately not used as predic tors as there were no differences for these variables between ICD a nd CAD patients. A seri es of hierarchical regression analyses were used to analyze the predictors of sleep efficiency. For the full sample of sleep diary data, EF %, distress, somatic hypervigilance, physical activity, ICD shock history, and mean sleep onset latency (sleep diar y) were used as predictors (Table 3-23). Mean waking after sleep onset (sleep diary) and total sleep time (sleep diary) were excluded as predictors due to multicollinearity with mean sleep efficiency in the CAD sample. This 6-predictor model was significant ( F (6,24) = 7.212, p = 0.005) at predicting 83% (71% adjusted ) of the variance in sleep efficiency scores. For the full sample of actigraphy data, EF%, distress, somatic hypervigilance, physical activity levels, ICD shock history, and mean sleep onset latency (ac tigraphy) scores were used as predictors (Table 3-24). Again, mean waking af ter sleep onset (actigraphy ) and total sleep time (actigraphy) were excluded as predictors due to multicollinearity with mean sleep efficiency in the CAD sample. This 6-predictor model was not significant at predic ting variance in sleep efficiency scores. A third regression analysis was conducted to see how well the predictors functioned for sleep diary data in the CAD sample (Table 3-25). The predictors in this 4-factor model included 50

PAGE 51

distress, somatic hypervigilance, physical activity scores, and m ean sleep onset latency (sleep diary). This model was marginally significant ( F (4,11) = 4.604, p = .062) at predicting 79% (62% adjusted) of the variance in sleep efficiency scores. A fourth regression analysis was conducted to an alyze predictors of sleep diary data in the ICD sample (Table 3-26). Predictors in this 6-factor model included EF%, distress, somatic hypervigilance, physical ac tivity levels, shock history, and mean sleep onset latency. This model was significant (F (6,8) = 5.332, p = .043) at predicting 87% (70% adjusted) of the variance in sleep efficiency. A fifth regression analysis was conducted to analyze predictors of sleep efficiency for actigraphy in the CAD sample (Table 3-27). Predictors in this 5-factor model included distress, somatic hypervigilance, physical activity, mean sleep onset latency (actigraphy), and mean waking after sleep onset (actigraphy). The mode l was not significant at predicting variance in sleep efficiency scores. Modeling with actigra phy data for ICD patients could not be conducted due to multicollinearity of predictors, toleran ce approaching zero, and artificial inflation of variance predicted due to a high number of predictors given the studys sample size. Multiple Regression Models A second round of regression analyses was conducted in order to examine robust predictors of sleep efficiency in the full, actigra phy, and sleep diary sample s. The rationale here was to decrease the number of predictors, em ploying only those that correlated with sleep efficiency at r = 0.4 0.7. Employing only strongly correlated predictors while simultaneously excluding predictors that could lead to multicolli nearity could result in more accurate prediction of the variance in sleep efficiency scores. For the full sample of actigraphy data, physical activity (DASI) scores and EF% were used as predictors that were entered simultaneously (see Table 3-28). Physical activity and EF% 51

PAGE 52

significantly predicted 56% (49% adjusted) of the variance in sleep efficiency scores, F (2,26) = 8.322, p = .005. Both EF% (p = .025) and physical activity scores ( p = .039) served as unique predictors in this model. For ICD patients in the sleep diary only ar m, physical activity (DASI) scores, somatic hypervigilance, and mean sleep onset latency (sleep di ary) were used as predictors (see Table 3-29). These variables signi ficantly predicted 78% (70% adjusted) of the variance in sleep efficiency scores, F (3,27) = 9.554, p = .005. Other follow-up multiple regressions were not completed due to a lack of corresponding variables that significantly correlated with sleep efficiency in the r = 0.4 0.7 range (see Tables 7 11). 52

PAGE 53

Base Day 1: Completed Psychosocial Self-Report Measures ( N = 59) Day 1: Patients Meeting Study Criteria Gave Informed Consent ( N = 64) Assignment to Sleep Diary arm or Slee p Diar y + Acti g ra p h y ar m Sleep Diary Only arm (n = 31): Days 1-14: Sleep Diary Data Collection ( n = 16 CAD patients and n = 15 ICD patients) Sleep Diary + Actigraphy arm (n = 29): Days 1-14 : Sleep Diary plus Actigraphy Data Collection ( n = 14 CAD patients and n = 15 ICD patients) Stud y Completion ( N = 60) Figure 3-1. Sample derivation and data collection process 53

PAGE 54

Table 3-1. Descriptive statisti cs on demographic, medical, slee p, and psychosocial variables for overall sample Variable n Mean/% SD Minimum Maximum Demographic variables Age 59 66.90 10.18 40 86 Sex (Male) 59 61.5% Race Caucasian 53 81.5% African-American 5 7.7% Asian/Pacific Islander 1 1.5% Marital status Single 1 1.5% Separated/Divorced 5 7.7% Widowed 3 4.6% Married/Remarried 42 64.6% Living with partner 4 6.2% Have children (Yes) 55 93.2% Number of children 55 2.64 1.40 0 7 Religion Catholic 9 13.8% Jewish 2 3.1% Protestant 42 64.6% Other 5 7.7% Employment status Full-time 10 15.3% Homemaker 2 3.1% Unemployed 3 4.6% Disability/Financial asst. 12 18.5% Retired 32 49.2% Income <$14,000 4 6.2% $15,000 29,999 9 13.8% $30,000 44,999 8 12.3% $45,000 59,999 7 10.8% $60,000 74,999 7 10.8% $75,000 89,999 12 18.5% 54

PAGE 55

Table 3-1. Continued Variable n Mean/% SD Minimum Maximum Sleep variables Actigraphy: Sleep efficiency 28 76.75% 10.56% 50.13% 91.82% Sleep onset latency 28 30.14 19.02 2.73 73.69 Waking after sleep onset 28 55.86 27.73 18.38 138.77 No. of awakenings 28 32.57 11.02 18.31 61.31 Total sleep time 28 386.80 77.87 210.00 498.23 Sleep Diary: Sleep efficiency 29 83.09% 9.36% 53.80% 98.59% Sleep onset latency 29 32.66 29.51 1.93 135.00 Waking after sleep onset 29 30.54 26.83 1.00 119.79 No. of awakenings 29 1.86 0.89 0.70 4.07 Total sleep time 29 416.22 75.95 3.84 8.83 Sleep quality (diary) 29 3.47 0.66 2.14 5 Sleep quality (PSQI total score) 29 7.71 4.87 0 19 Daytime Dysfunction (PSQI item #8b) 29 0.91 1.11 1 3 Psychosocial variables Distress (HADS) 57 10.68 7.99 1.00 39.00 Hypervigil. (BVS) 52 21.17 9.41 .00 40.00 Phy. activity (DASI) 57 25.44 12.81 4.50 58.20 Shock Anxiety 29 17.31 8.56 9.00 42.00 55

PAGE 56

Table 3-2. Means and standard deviations/frequ encies for medical, sleep, and psychosocial variables by cardiac condition Variable ICD Patients CAD Patients Medical variables ICD type: Single 40.0% (n = 12) --Dual 36.7% (n = 11) --Biventricular 23.3% (n = 7) --No. of shocks, previous two years 1.84 + 4.76 --No. participants who received at least one shock, previous 2 years 9 (30%) --Months since ICD implantation 53.57 + 45.47 (4.5 years), min = 3; max = 168 --EF% 35.53% + 13.12% 48.69% + 12.58% Have CAD (Yes) 58.1% 100% Hx of MI (Yes) 29.0% 44.8% Hx of CABG (Yes) 38.7% 41.4% Angina (SAQ): Physical limitation --52.44 + 18.31 Anginal stability --63.79 + 27.21 Anginal frequency --81.39 + 21.17 Treatment satisfact. --89.45 + 19.15 Disease perception --60.06 + 21.52 Sleep variables Actigraphy: Sleep efficiency 82.80% + 6.39% 69.76% + 10.24% Sleep onset latency 25.25 + 16.04 35.79 + 21.20 Waking after sleep onset 45.82 + 22.50 67.43 + 29.48 No. of awakenings 31.09 + 10.96 34.29 + 11.28 Total sleep time 7.18 + .72 hours (430.65 + 43.04 minutes) 5.60 + .12 hours (336.19 + 79.39 minutes) Sleep Diary: Sleep efficiency 82.26% + 8.39% 83.95% + 10.35% Sleep onset latency 33.83 + 27.65 31.40 + 31.84 Waking after sleep onset 32.56 + 26.38 28.46 + 27.62 No. of awakenings 2.07 + .84 1.63 + 0.74 Total sleep time 6.90 + 1.15 hours (414.03 + .02 minutes) 6.98 + 1.12 hours (418.64 + .02 minutes) Sleep quality (diary) 3.47 + 0.33 3.46 + .67 56

PAGE 57

57 Table 3-2. Continued Variable ICD Patients CAD Patients Sleep quality (PSQI total scores) 8.06 + 4.75 7.32 + 5.06 Daytime Dysfunction (PSQI item #8b) 1.16 + 1.16 0.63 + 1.01 Psychosocial variables Distress (HADS) 10.72 + 7.50 10.64 + 8.60 Hypervigil. (BVS) 19.24 + 10.89 23.36 + 7.01 Phy. activity (DASI) 23.62 + 10.98 27.50 + 14.56 Shock Anxiety (FSAS) 17.31 + 8.56 --Physical Health (SF-12) 34.86 + 8.98 37.38 + 10.34 Mental Health (SF-12) 51.44 + 8.05 52.82 + 10.70 Table 3-3. Means and standard deviations for sleep variable s for sleep diary study arm and actigraphy data from sleep diary plus actigraphy arm Variable Sleep Diary Actigraphy Sleep efficiency 80.65 + 7.98 % 76.03 + 10.51% Sleep onset latency 32.66 + 29.51 30.14 + 19.02 Waking after sleep onset 30.54 + 26.83 55.86 + 27.73 Total sleep time 6.84 + 1.12 hours (410.69 + 67.40 minutes) 6.40 + 1.30 hours (384.29 + 78.19 minutes)

PAGE 58

Table 3-4. Pearsons product-moment co rrelations among psyc hosocial variables Variable 1 2 3 4 5 6 7 8 1 Depression and anxiety (HADS) --2 Hypervigilance (BVS) .27 --3 Physical activity (DASI) -.44 ** -.16 --4 Physical limitations (SAQ) .45 -.03 -.62 ** --5 Anginal stability (SAQ) .00 .14 -.08 .24 --6 Anginal frequency (SAQ) .50 ** .04 -.52 ** .28 .36 --7 Treatment satisfaction (SAQ) .58 ** .06 -.21 -.03 .43 .74 --8 Disease perception (SAQ) .54 .08 -.07 .37 .48 .35 .40 --*p < 0.05, 2-tailed; ** p < 0.01, 2-tailed 58

PAGE 59

Table 3-5. Pearsonss product-moment correlations among sleep variables Variable 1 2 3 4 5 6 7 8 9 10 11 12 1 Sleep efficiency (A) --2 Sleep onset latency (A) -.31 --3 Waking after sleep onset (A) -.64 ** -.02 --4 Total sleep time (A) .79 ** -.18 -.33 --5 No. of awakenings (A) -.26 -.03 .68 ** .06 --6 Sleep efficiency (SD) .48 -.24 -.13 .31 .17 --7 Sleep onset latency (SD) -.04 .36 -.04 -.04 -.06 -.59 ** --8 Waking after sleep onset (SD) -.09 -.08 -.06 -.01 -.14 -.67 ** .32 --9 Total sleep time (SD) .38 -.04 .05 .68 ** .41 .72 ** -.48 ** -.45 ** --10 No. of awakenings (SD) .15 -.26 .10 .19 .17 .041 -.20 .15 .16 --11 Sleep quality (SD) .37 -.44 -.03 .46 .12 .28 -.30 -.24 .35 -.17 --12 Sleep quality (PSQI) .09 -.02 -.16 -.01 -.48 -.42 ** .54 ** .38 ** -.56 ** -.22 -.34 --*p < 0.05, 2-tailed; ** p < 0.01, 2-tailed. A = actigraphy; SD = sleep diary; PSQI = Pittsburgh Sleep Quality Index. 59

PAGE 60

Table 3-6. Pearsons product-moment correl ations among sleep efficiency and correspondi ng predictors for full sample of actigr aphy indices Variable 1 2 3 4 5 6 7 8 9 1 Age --2 Gender -.05 --3 Ejection fraction % -.01 -.01 --4 Distress (HADS) -.11 .16 -.01 --5 Som. hypervigilance (BVS) .05 .01 .34 .22 --6 Physical activity (DASI) -.20 -.24 -.05 -.38 ** -.16 --7 Shock history .09 -.05 -.17 .17 .26 .17 --8 Mean SOL -.01 -.06 .28 -.05 .57 ** -.05 .04 --9 Mean SE -.08 .06 -.43 -.26 -.21 .53 .12 -.31 --*p < 0.05, 2-tailed; ** p < 0.01, 2-tailed. SOL = sleep onset latency; SE = sleep efficiency. 60

PAGE 61

Table 3-7. Pearsons product-moment correl ations among sleep efficiency and correspondi ng predictors for full sample of sleep diary indices Variable 1 2 3 4 5 6 7 8 9 1 Age --2 Gender .14 --3 Ejection fraction % .21 .01 --4 Distress (HADS) -.22 .15 -.33 --5 Som. hypervigilance (BVS) .19 .25 .34 -.03 --6 Physical activity (DASI) -.11 -.09 .16 -.28 .10 --7 Shock history .09 .01 -.17 .29 .29 -.01 --8 Mean SOL .05 .22 -.31 .28 -.34 -.13 .07 --9 Mean SE .03 .05 .11 -.10 .42 .09 .01 -.74 ** --*p < 0.05, 2-tailed; ** p < 0.01, 2-tailed. SOL = sleep onset latency; SE = sleep efficiency. 61

PAGE 62

Table 3-8. Pearsons product-moment corr elations among sleep efficiency and corr esponding predictors for CAD sample of actigraphy indices Variable 1 2 3 4 5 6 7 8 1 Age --2 Gender -.18 --3 Distress (HADS) -.10 .21 --4 Som. hypervigilance (BVS) -.19 .13 .07 --5 Physical activity (DASI) -.11 -.41 -.41 .12 --6 Mean SOL .23 .00 -.30 .21 -.00 --7 Mean WASO -.57 .07 .07 -.09 -.13 -.24 --8 Mean SE .03 -.20 -.41 .05 .60 -.08 -.56 --*p < 0.05, 2-tailed; ** p < 0.01, 2-tailed. SOL = sleep onset latency; WASO = waking after sleep onset; SE = sleep efficiency. 62

PAGE 63

Table 3-9. Pearsons product-moment correl ations among sleep efficiency and correspondi ng predictors for CAD sample of sleep diary indices Variable 1 2 3 4 5 6 7 1 Age --2 Gender -.23 --3 Distress (HADS) -.20 .39 --4 Som. hypervigilance (BVS) -.22 .34 .20 --5 Physical activity (DASI) -.24 -.12 -.38 .15 --6 Mean SOL .50 .47 .28 -.17 -.41 --7 Mean SE -.48 -.33 -.15 -.03 .38 -.82 ** --*p < 0.05, 2-tailed; ** p < 0.01, 2-tailed. SOL = sleep onset latency; SE = sleep efficiency. 63

PAGE 64

Table 3-10. Pearsons product-moment corr elations among sleep efficiency and corr esponding predictors for ICD sample of actigraphy indices Variable 1 2 3 4 5 6 7 8 9 10 1 Age --2 Gender .05 --3 Ejection fraction % -.11 .10 --4 Distress (HADS) -.12 .10 .08 --5 Som. hypervigilance (BVS) .09 -.04 .38 .34 --6 Physical activity (DASI) -.34 -.06 -.06 -.37 -.47 --7 Shock history .16 -.09 -.09 .27 .39 -.01 --8 Mean SOL -.23 -.04 .21 .32 .72 -.41 .24 --9 Mean WASO -.16 .10 -.32 .10 -.24 .28 -.28 -.00 --10 Mean SE -.15 .09 .32 -.09 -.14 .43 -.27 -.37 -.54 --*p < 0.05, 2-tailed; ** p < 0.01, 2-tailed. SOL = sleep onset latency; WASO = waking after sleep onset; SE = sleep efficiency. 64

PAGE 65

65 Table 3-11. Pearsons product-moment correl ations among sleep efficiency and correspondi ng predictors for ICD sample of sleep diary indices Variable 1 2 3 4 5 6 7 8 9 1 Age --2 Gender .51 --3 Ejection fraction % -.06 .15 --4 Distress (HADS) -.20 -.10 -.12 --5 Som. hypervigilance (BVS) .46 .06 .27 -.19 --6 Physical activity (DASI) -.13 -.19 .00 -.12 -.14 --7 Shock history .21 .06 -.11 .40 .50 .18 --8 Mean SOL -.43 -.16 -.25 .26 -.55 .43 .09 --9 Mean SE .37 .21 .15 .00 .77 -.50 .10 -.67 ** --*p < 0.05, 2-tailed; ** p < 0.01, 2-tailed. SOL = sleep onset latency; SE = sleep efficiency.

PAGE 66

Table 3-12. ANCOVA results for mean sleep e fficiency (sleep diar y) by cardiac condition Source Sum of squares df Mean square F p Effect size Observed power Corrected model 344.228 3 114.743 1.711 .180 .111 .414 Intercept 24612.261 1 24612.261 367.025 .000 .900 1.000 Cardiac condition 283.364 1 283.364 4.226 .046 .093 .519 EF% 189.088 1 189.088 2.820 .101 .064 .375 Cardiac condition *EF% 208.597 1 208.597 3.111 .085 .071 .406 Error 2749.414 41 67.059 Total 312668.350 45 Corrected total 3093.642 44 R Squared = .111 (Adjusted R Squared = .046) Table 3-13. ANCOVA results for mean number of awakenings (sleep diary) by cardiac condition Source Sum of squares df Mean square F p Effect size Observed power Corrected model .851 3 .284 .395 .757 .028 .121 Intercept 9.303 1 9.303 12.950 .001 .240 .940 Cardiac condition .026 1 .026 .037 .849 .001 .054 EF% .151 1 .151 .211 .649 .005 .073 Cardiac condition *EF% .181 1 .181 .251 .619 .006 .078 Error 29.453 41 .718 Total 206.983 45 Corrected total 30.305 44 R Squared = .028 (Adjusted R Squared = -.043) 66

PAGE 67

Table 3-14. ANCOVA results for mean sleep quality (sleep diar y) by cardiac condition Source Sum of squares df Mean square F p Effect size Observed power Corrected model .366 3 .122 .271 .846 .019 .097 Intercept 32.234 1 32.234 71.422 .000 .630 1.000 Cardiac condition .100 1 .100 .220 .641 .005 .074 EF% .053 1 .053 .118 .733 .003 .063 Cardiac Condition *EF% .022 1 .022 .049 .826 .001 .055 Error 18.955 42 .451 Total 568.020 46 Corrected total 19.321 45 R Squared = .019 (Adjusted R Squared = -.051) Table 3-15. ANCOVA results for mean sleep on set latency (actigraphy) by cardiac condition Source Sum of squares df Mean square F p Effect size Observed power Corrected model 1116.236 3 372.079 .787 .517 .116 .185 Intercept 370.184 1 370.184 .783 .388 .042 .134 Cardiac condition 59.799 1 59.799 .127 .726 .007 .063 EF% 44.318 1 44.318 .094 .763 .005 .060 Cardiac condition *EF% 15.515 1 15.515 .033 .858 .002 .053 Error 8508.487 18 472.694 Total 29746.419 22 Corrected total 9624.723 21 R Squared = .116 (Adjusted R Squared = -.031) 67

PAGE 68

Table 3-16. ANCOVA results for mean number of awakenings (actigraphy) by cardiac condition Source Sum of squares df Mean square F p Effect size Observed power Corrected model 384.785 3 128.262 .883 .468 .128 .204 Intercept 621.371 1 621.371 4.279 .053 .192 .499 Cardiac condition 50.141 1 50.141 .345 .564 .019 .086 EF% .812 1 .812 .006 .941 .000 .051 Cardiac condition *EF% 113.414 1 113.414 .781 .388 .042 .133 Error 2613.769 18 145.209 Total 26742.534 22 Corrected total 2998.554 21 R Squared = .128 (Adjusted R Squared = -.017) Table 3-17. ANOVA results for mean sleep e fficiency (actigraphy) by cardiac condition Source Sum of squares df Mean square F p Effect size Observed power Corrected model 1184.237 1 1184.237 16.840 .000 .393 .977 Intercept 162087.590 1 162087.590 2304.881 .000 .989 1.000 Cardiac condition 1184.237 1 1184.237 16.840 .000 .393 .977 Error 1828.414 26 70.324 Total 167926.912 28 Corrected total 3012.651 27 R Squared = .393 (Adjusted R Squared = .370) 68

PAGE 69

Table 3-18. ANOVA results for mean total sleep time (actigraphy) by cardiac condition Source Sum of squares df Mean square F p Effect size Observed power Corrected model 62145.437 1 62145.437 15.908 .000 .380 .970 Intercept 4095346.267 1 4095346.237 1048.36 .000 .976 1.000 Cardiac condition 62145.437 1 62145.437 15.908 .000 .380 .970 Error 101567.217 26 3906.431 Total 4352818.542 28 Corrected total 163712.653 27 R Squared = .380 (Adjusted R Squared = .356) Table 3-19. ANOVA results for mean waking after sleep onset (actigraph y) by cardiac condition Source Sum of squares df Mean square F p Effect size Observed power Corrected model 3252.621 1 3252.621 4.828 .037 .157 .562 Intercept 89331.155 1 89331.155 132.605 .000 .836 1.000 Cardiac condition 3252.621 1 3252.621 4.828 .037 .157 .562 Error 17515.284 26 673.665 Total 108126.242 28 Corrected total 20767.905 27 R Squared = .157 (Adjusted R Squared = .124) 69

PAGE 70

Table 3-20. ANOVA results for mean total sl eep time (sleep diary) by cardiac condition Source Sum of squares df Mean square F p Effect size Observed power Corrected model .086 1 .086 .067 .797 .001 .057 Intercept 5789.314 1 2789.314 2168.95 .000 .975 1.000 Cardiac condition .086 1 .086 .067 .797 .001 .057 Error 72.017 56 1.286 Total 2863.671 58 Corrected total 72.103 57 R Squared = .001 (Adjusted R Squared = -.017) Table 3-21. Kruskal-Wallis ANOVA results for mean sleep onset latency (sleep diary) by cardiac condition Source Chi-Square df p Mean rank (ICD) Mean rank (CAD) Corrected model .787 1 .375 31.40 27.46 Table 3-22. Kruskal-Wallis ANOVA re sults for mean waking after sl eep onset (sleep diary) by cardiac condition Source Chi-Square df p Mean rank (ICD) Mean rank (CAD) Corrected model .757 1 .384 30.88 27.05 70

PAGE 71

Table 3-23. Summary of hierarchical regression analysis fo r predictors of sleep diary sleep efficiency for overall sample Variable B SE B t p R 2 R 2 Step one EF% -.165 .108 -.289 -1.525 .144 .004 .004 Step two HADS .262 .311 .194 .840 .411 BVS .337 .177 .411 1.900 .073 .273 .269 Step three DASI .117 .154 .177 .764 .454 .273 .000 Step four ICD shock history -.196 .304 -.125 -.643 .528 Sleep diary SOL -.209 .097 -.443 -2.150 .045 .436 .163 SOL = sleep onset latency 71

PAGE 72

Table 3-24. Summary of hierarchical regression analysis for predictors of actigraphy sleep efficiency for overall sample Variable B SE B t p R 2 R 2 Step one EF% -.257 .446 -.291 -.576 .623 .566 .566 Step two HADS -1.297 1.415 -.832 -.916 .456 BVS .070 .485 .072 .145 .898 .741 .175 Step three DASI -.038 .489 -.055 -.077 .946 .744 .003 Step four ICD shock history 1.530 3.664 .148 .418 .717 .750 .006 Step five Actigraphy SOL .176 .220 .332 .803 .506 .81 .061 SOL = sleep onset latency 72

PAGE 73

Table 3-25. Summary of hierarchical regression analysis fo r predictors of sleep diary sleep efficiency for CAD sample Variable B SE B t p R 2 R 2 Step one BVS -.470 .306 -.361 -1.537 .185 HADS .792 .503 .407 1.576 .176 .008 .008 Step two DASI .127 .214 .144 .593 .579 .110 .103 Step three Sleep diary SOL -.330 .083 -.940 -3.979 .011 .786 .676 SOL = sleep onset latency 73

PAGE 74

Table 3-26. Summary of hierarchical regression analysis fo r predictors of sleep diary sleep efficiency using ICD sample Variable B SE B t p R 2 R 2 Step one EF% -.152 .176 -.161 -.860 .429 .000 .000 Step two HADS .668 .481 .353 1.390 .223 BVS 1.118 .360 .974 3.106 .027 .640 .639 Step three DASI -.071 .264 -.060 -.269 .799 .728 .089 Step four ICD shock history -.695 .428 -.471 -1.623 .165 .834 .106 Step five Sleep diary SOL -.125 .116 -.233 -1.070 .334 .865 .031 SOL = sleep onset latency 74

PAGE 75

Table 3-27. Summary of hierarchical regression analysis for predictors of actigraphy sleep efficiency using CAD sample Variable B SE B t p R 2 R 2 Step one HADS -.213 .423 -.181 -.504 .649 BVS -.016 .514 -.009 -.032 .976 .237 .237 Step two DASI .257 .171 .462 1.501 .230 .411 .174 Step three Actigraphy SOL .051 .239 .096 .214 .844 Actigraphy WASO -.206 .168 -.564 -1.228 .307 .757 .346 SOL = sleep onset latency; WA SO = waking after sleep onset 75

PAGE 76

Table 3-28. Summary of multiple regression analysis for predictors of actigraphy sleep efficiency using full sample Variable B SE B t p R 2 R 2 DASI .276 .120 .442 2.298 .039 EF % -.374 .148 -.487 -2.528 .025 .561 .561 Table 3-29. Summary of multiple regression analysis for pred ictors of sleep diary sleep efficiency using ICD sample Variable B SE B t p R 2 R 2 DASI -.352 .204 -.297 -1.725 .123 BVS .636 .226 .555 2.812 .023 Sleep diary SOL -.168 .109 -.315 -1.547 .160 .782 .782 SOL = sleep onset latency 76

PAGE 77

CHAPTER 4 DISCUSSION The purpose of this study was to compare sleep patterns between CAD and ICD patients using both objective and subjective sleep data. A second major aim was to analyze predictors of adaptive sleep efficiency in the overall sample. This study is the first study to date to use actigraphy in objectively measuring sleep patte rns in electrophysiology patients. Actigraphy measures were compared to subjective sleep diary data in order to compare these two patterns. The current study is also one of the first to co mpare sleep patterns among patients with different cardiac diagnoses. Thus, this study provided a seri es of comparisons among subjective versus objective sleep measures and coronary disease versus ICD patients. There were two major findings from this study. First, CAD patients had poorer sleep compared to ICD patients in terms of sleep efficiency and total sleep time. Second, several psychological and medical variables significantly predicted variance in sleep efficiency scor es. Depending on the model employed, these variables included somatic hypervigilance, psychological distress, physical activity level, sleep onset latency, and EF%. Patient Characteristics In examining patient characteristics, this full sample is similar to other cardiac populations in terms of demogr aphic variables. The modal patient was a 59-year-old married Caucasian male either working full-time or on disability. In terms of medications, most patients were on some combination of beta-blo ckers, calcium channel blockers, diuretics, cholesterol medication, and anti-a nginal medication. Beta-blocker diuretic, and gatroesophageal reflux medications are notorious for disrupting sleep. There were no differences found for betablocker and GERD medications between CAD an d ICD patients, although ICD patients were more likely to be on diuretic medication. Generally between 1/3 and 1/ 2 of ICD patients are 77

PAGE 78

shocked within their first year pos t-implantation (Sears, Shea, & Con ti, 2005). This is consistent with the current study in that 30% of participants had been shocked in th e two years previous to study entry. In examining the samples scores on psychologica l variables, one of ma in findings is that these individuals reported signi ficantly lower levels of depre ssion and anxiety than what is normally seen in the cardiac populati on. For depression, 7.1% of part icipants report ed clinically significant depression and 11.9% of participants reported mild depressive symptoms. These rates are much lower than those found by Sear s and colleagues at 2433% for clinically significant depressive symptoms in ICD patient s (Sears et al., 1999; Sears & Conti, 2006). Regarding normative rates of depression in C AD patients, numerous studies by Frasure-Smith and colleagues have found higher prevalence ra tes ranging from 31.5 41.4% (Lesperance, Frasure-Smith, & Talajic, 1996; Lesperance, Frasure-Smith, & Theroux, 2000). A recent American Heart Association (AHA) Science Advi sory (2008) concluded th at 15-20% of patients hospitalized for myocardial infarction meet DSM -IV criteria for major depressive disorder. A recent National Health Interview Survey ( 2007) found the 12-month prevalence of major depressive disorder to be 9.3% in a in a sample of 30,801 participants. Additionally, Glassman and Shapiro (1998) found that one in three individuals who have survived a heart attack are depressed. This discrepancy in depression rates between what was found in the current sample and rates found in the literature is not easily explained. The sample drawn tended to be an experienced heart disease group of patients with over 4 years with an ICD on average. Perhaps, the acute demands of coping with heart disease are more managed at this point. The differences may be also partially attributable to use of the HADS to measure this construct. 78

PAGE 79

While the HADS has excellent psychometric properties, it may have not been adequately sensitive to depressive symptoms that are como rbid with medical diseases such as cardiac disease. The scales have been widely used in inpatient settings and in European research but additional attention to its utility in US outpatient samples may be warranted. For anxiety, 9.8% of participants reported c linically significant a nxiety and 14.6% reported clinically significant anxiety symptoms according to the HADS. Sears and colleagues (1999) review found that 13-38% of ICD patients experi ence clinically significant anxiety and that anxiety symptoms commonly relate to shock anxiety and fears of device malfunction and death. In CAD patients, Framingham Offs pring study researchers found that anxiety is an independent risk factor for total mortality in men (Eaker et al., 2005). However, th e current study finding of a low mean level of anxiety among participants is more consistent with the work of Suls and Bunde (2005). Their qualitative review of 17 pu blished studies found that only four of these studies revealed a positive association between anxiety at baseline a nd subsequent cardiac morbidity or mortality, suggesting a limited link between anxiety and coronary disease. When sleep diary and actigraphy methods were compared regard less of cardiac diagnosis, participants had higher total sleep time and, by extension, higher sl eep efficiencies according to sleep diaries versus actigraphy. These findings are surprising in light of the fact that actigraphy usually overscores sleep, e.g., it sometimes misses the fact that a person is awake lying in bed but not moving their body. Lichstein and colleag ues (2005) review of validating actigraphy with insomnia patients found that actigra phy overscored sleep by 25-49 minutes depending on the scoring algorithm used. However, the current findings make sense when viewed in the context that people often wake up but do not remember these awakenings the next morning when recording the previous nights slee p in their sleep diary. For exampl e, if a participant woke three 79

PAGE 80

times in a given night, they may only remember one or two of their awakenings. The finding that the sleep diary method of assessing sleep actu ally overscored sleep mu st be interpreted in the context that this study included both good an d poor sleepers. Moreover, it may be that comparisons of general findings from sleep rese arch may be less generalizable. Many of the sleep studies examine individuals with self-repor ted problematic sleep, whereas participants in cardiac studies may be but are not necessa rily concerned about their sleep. CAD and ICD Participant Differences in Sleep The first aim of this study was to compare sl eep indices for CAD and ICD participants. It was hypothesized that ICD patients would have poorer sleep indice s related to hypervigilance for device functioning and nocturnal shock discharge. All three of the previous studies using actigraphy with cardiac patients were conducte d by Redeker et al. (1994, 1995, and 1999). This body of work has focused on women recovering from coronary artery bypass graft surgery. Redeker and colleagues have found that stronger circadian patterns of activ ity and increases in levels of activity were positively related to be tter physical function (recovery) and length of hospital stay. As these actigra phy studies focused exclusively on CAD patients, no comparisons were made with other cardiac samples. The current study found that according to actigraphy, CAD patients had significantly shorter total sleep times and by extension lower sleep efficiencies than ICD patients. This is in direct contrast to the hypothe sis that ICD patients would sleep worse due to hyperarousal. Several explanations were explored in order to account for this fi nding, including level of depression and anxiety reporte d in the HADS, somatic hypervig ilance reported in the BVS, physical activity level reported in the DASI, and EF%. Experiencing depression, physical symptoms of hypervigilance, or reporting a lowe r level of physical activity could account for CAD patients having worse sleep. However, di fferences between CAD and ICD patients for 80

PAGE 81

depression/anxiety, hypervigilance, a nd physical activity levels were not statisticall y significant. Also of note, CAD patients had expectably hige r EFs%s than ICD patients, such that poor perfusion of blood to the body cannot explain this finding, either. The finding that CAD patients had poorer sleep patterns than ICD patients highlights the potential importance of chest pain as a variable that affects sl eep. Nowlin et al. (1965) found that nocturnal angina occurred predominantly dur ing REM sleep and was associated with heart rate acceleration. When dream content could be reported, it included awareness of chest pain and emotions of fear, anger, and frustration. In the current study, physical limitations due to chest pain (e.g., activities of daily living and performing specific tasks), disease perception (degree to which chest pain interferes with lif e), and anginal stability over the previous four weeks were more problematic than the actual fre quency of chest pain and treatment satisfaction. In the ICD sample, 58.1% of participants had underlying coronary disease. Thus, 80% of the total sample had underlying coronary disease, which could lead to angi na. Unfortunately, the SAQ was administered only to participants with C AD without an ICD. It would be interesting to compare chest pain across the two subgroups in or der to measure potential differences. Future studies in this area could bene fit from administering the SAQ across the board in order to thoroughly measure and make comp arisons for this construct. CAD patients differ widely in terms of the num ber of vessels occluded and the degree of occlusion present in each of these vessels. The medical severi ty of coronary disease may or may not correlate with the severity of complaints for physical symptoms such as chest pain. However, it is possible that ICD patients had grea ter disease severity regarding coronary disease than the CAD patients themselves. Future analys es for the current study w ill take these variables into account. 81

PAGE 82

The finding that CAD patients had poorer sleep patterns than ICD patients may be attributed to another more speculative explanatio n. Actigraphy is a highly practical measure of objective sleep to use in research due to its affordability, convenie nce, and ability to assess sleep in the natural sleeping environment. Actig raphy has been validated with polysomnography, which makes it a strong choice. However, both of these measures use gross measures to assess whether the patient is asleep or awake in a give n epoch of sleep. Both electroencephalographic measures for polysomnography and movement of the extremities for actigraphy represent gross measures. In fact, localized areas of the brain ma y still be awake at a point in time when gross measures suggest that the individual is asleep (Alam, McGinty, Bashir, Kumar, & Imeri et al., 2004; Yasuda, Yoshida, Garcia-Garcia, Kay, & Kr ueger, 2004). Patterns in localized brain activity that occur differentially across CAD and ICD patients could potentially explain the finding that CAD patients have poorer patterns than ICD patients. Predictors of Sleep Efficiency in the Full, CAD, and ICD Samples The second aim of this study was to analyze predictors of a cruc ial sleep index, sleep efficiency, in cardiac patients. It was hypothesized that age, sex, ejection fraction, psychological distress, somatic hypervigilance, and daytime phys ical activity levels wo uld strongly predict sleep efficiency in the overall study sample. While a series of regression analyses were conducted, key patterns emerged from the significan t results. First, daytime physical activity level per DASI scores was an important predictor. Physical exercise fa cilitates sleep, although it is difficult for many cardiac patients to engage in this regularly due to ischemia during vigorous activity. Second, hypervigilance per BVS scores em erged as a common significant predictor. Somatic and cognitive hypervigilance inhibit slee p, and many of the symptoms of having cardiac disease overlap with symptoms of hypervigilance. Increasingly, insomnia is being conceptualized as a disorder of free-floating a nd generalized hypervigilanc e that is present both 82

PAGE 83

day and night (Stepanski, Zorick, Roehrs Young, & Roth, 1988; Lichstein, Wilson, Noe, Aguillard, & Bellur, 1994; Varkevisser, Van Dongen, & Kerkhof, 2005). The current study included both good sleepers and poor sleepers. Future work in this area could focus on recruiting only participants with insomnia, as this could provide furthe r evidence of 24-hour hypervigilance in insomnia. Administration of a cognitive measure of hypervigilance is warranted for future work. Limitations The current study has several methodologica l limitations that may require cautious interpretation of findings. When using retros pective self-report quest ionnaires, inaccurate memory and selective recall biases may operate. Some patients with insomnia have a natural tendency to focus on the worst experiences and perhaps to amplify their importance (Spielman, Yang, & Glovinsky, 2005). This dynamic may have resulted in some participants reporting longer sleep latencies, or more or longer wakings after sleep onse t than what actu ally occurred. The current study used actigraphy with one-half of the CAD participan ts and one-half of the ICD participants due to logi stic constraints. Using actigraphy for the entire sample cohort instead of a subset of patients would aid the interpretation of findings by increasing statistical power. It would also be of use to recruit a la rger sample size overall, as this would allow for more robust multivariate data analysis. The findings from this study may not necessarily generalize to patients with other forms of comorbid sleep disturbance, such as patients wi th cancer or chronic pain experiencing sleep problems. Also, this study excluded patients with obstructive sleep apnea, so these data may not generalize to that population. Additionally, a limitation of this study involves the use of sleep onset latency (for both sleep diary and actigraphy measures) in the predic tion of sleep efficienc y. Sleep onset latency 83

PAGE 84

naturally relates to the calculation of sleep effici ency in any sample of sleep data. Thus, this variable should not be included in a prediction model and will not be used in future analyses. A final limitation of the current study is the inclusion of some participants with cardiac diagnoses or conditions in addition to those unde r study, CAD and ICDs. In this sample, there were participants with AF (13 participants, 22% ), pacemakers (4 participants, 7%), congestive heart failure (2 participants, 3%), and cardiomyopathy (1 participant, 2%). Logistical constraints barred the derivation of a completely pure cardiac sample. Significance The proposed study furthers our understanding of how sleep patterns relate to cardiac disease. Few researchers have previously examined the sleep-cardiac disease relationship, and existing studies have examined single outcomes su ch as total sleep time and sleep quality. The current study specifically focused on how CA D and ICD patients may differ on a broad, thorough range of sleep indices and to what we c ould attribute any potential differences. Sleep efficiency is an important index of sleep, as the range of values that i ndicate adaptive sleep for this index and interindividual variability for this index are smaller than in sleep indices such as total sleep time. Sleep efficiency is also a useful outcome measure in that it is the variable used to titrate some robust behavioral sleep interven tions, such as sleep restriction (Perlis, 2005). This study also yielded valuable information by documenting the feasibility of using sleep diaries and actigraphy as reliable da ta collection methods in cardiac patients. The literature lacks studies that have examined usi ng sleep diaries in cardiac patients, despite the fact that this method of data collection correlates highly with objective sleep indices, is inexpensive, and is practical for most researchers. This study is th e only study to use actigra phy in measuring sleep in electrophysiology patients to date, and only the fourth stud y to use actigraphy with CAD 84

PAGE 85

patients. Given the degree of poor sleep in bot h the ICD and CAD samples in this study, sleep should be considered a salient aspect of the disease process in cardiology. Conclusions Two important conclusions emerged from the current study. The first of these findings concerns the comparison of sleep between IC D and CAD patients. While it was hypothesized that ICD patients would have poorer sleep related to a nxiety surrounding possible shock discharge, it was CAD patients that had poorer sleep per actigraphy measures. This finding highlights the potential importance of chest pain in factors that may infl uence sleep patterns in cardiac patients. The second of these findings concerns predictors of sleep efficiency in cardiac patients. Depending on the type of regression model em ployed, somatic hypervigilance, psychological distress, physical activity level, sleep onset latency, and EF% were significant predictors of sleep efficiency in this sample. Some of these vari ables are modifiable lifestyle factors, emphasizing the potential utility of psychoeducation and treatment for cardiac patie nts with sleep disturbance. Cognitive Behavioral Therapy for Insomnia (CBT-I) is an empirically validated treatment that has not yet been investigated fo r use with cardiac patients. So matic hypervigilance is a variable that overlaps significantly with both cardiac disease and insomnia suggesting that use of CBT-I with cardiac patients should include practicing the relaxation response. 85

PAGE 86

LIST OF REFERENCES Alam, N., McGinty, D., Bashir, T ., Kumar, S.,& Imeri, L. et al. (2004). Interleukin-1 modulates state-dependent discharge activity of preoptic area and basal forebrain neurons: Role in sleep regulation. European Journal of Neuroscience, 20 207-216. American Academy of Sleep Medicine, Standards of Practice Committee of the American Academy of Sleep Medicine. (2003). Practice parameters for using Polysomnography to evaluate insomnia: An update. Sleep, 26 754-760. American Heart Association (2006). Heart and Stroke Statistical Update Dallas, TX: American Heart Association. American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4 th ed., text revision). Washington, DC: American Psychiatric Press. American Psychological Association (2002). Ethical principles of psychologists and code of conduct. American Psychologist, 57 1060-1073. American Sleep Disorders Associati on Standards of Practice Committee, Polysomnography Task Force. (1997). Practice parameters for the indications for polysomnography and related procedures. Sleep, 20 406-422. American Sleep Disorders Association (1990). International classification of sleep disorders: Diagnostic and coding manual. Rochester, MN: American Sleep Disorders Association. Ancoli-Israel, S., Cole, R., Alessi, C., Chambe rs, M., Moorcroft, W., & Pollack, C.P. (2003). The role of actigraphy in the st udy of sleep and circadian rhythms. Sleep, 26, 342-392. Ancoli-Israel, S. (2005). Actigraphy. In M.H. Kryger & W.C. Dement (Eds.), Principlesand Practice of Sleep Medicine Philadelphia: Saunders. Ayas, N.T., White, D.P., Manson, J.E., Stampf er, M.J., Speizer, F.E., et al. (2003). A prospective study of sleep duration a nd coronary artery disease in women. Archives of Internal Medicine, 163 205 209. Bardy, G.H. Lee, K.L., Mark, D.B., Poole, J.E., Packer, D.L. Boineau, R., & Sudden Cardiac Death in Heart Failure Trial (SCD-HeFT) Investigators. (2005). Amiodarone or an implantable cardiove rter-defibrillator for congestive heart failure. New England Journal of Medicine, 352 225-237. Bhatia, S.C., & Bhatia, S.K. (1999). Depression in women: diagnostic and treatment considerations. American Family Physician, 60, 225-240. 86

PAGE 87

Benca, R.H. (2005). Mood disorders. In M.H. Kryger & W.C. Dement (Eds.), Principles and Practice of Sleep Medicine Philadelphia: Saunders. Berry, R.B. (2003). Sleep Medicine Pearls (2 nd Ed.). Philadelphia: Hanley & Belfus. Blood, M.L., Sack, R.L., Percy, D.C., & Pen, J.C. (1997). A comparison of sleep Detection by wrist actigraphy, beha vioral response, and polysomnography. Sleep, 20, 388-395. Bjelland, I., Dahl, A.A., Haug, T.T., & Neckelmann, D. (2002). The validity of the Hospital Anxiety and Depression Scale. An updated literature review. Journal of Psychosomatic Research, 52, 69-77. Braunwald, E. (1992). Assessment of car diac function. In E. Braunwald (Ed.), Heart Disease: A Textbook of Cardiovascular Medicine Philadelphia: Saunders. Breslau, N., Roth, T., Rosenthal, L., & Andr eski, P. (1996). Sl eep disturbance and psychiatric disorders: A longitudinal epidemiol ogical study of young adults. Biological Psychiatry, 39 411-418. Buysse, D.J., Reynolds, C.F., Monk, T.H., Berman, S.R., & Kupfer, D.J. (1989). The Pittsburgh Sleep Quality Index: A new instrument for psychiatric practice and research. Psychiatry Research, 28, 193-213. Buysse, D.J., Ancoli-Israel, S., Edinger, J. D., Lichstein, K.L., & Morin, C.M. (2006). Recommendations for a standard re search assessment on insomnia. Sleep, 29 1155-1173 Buysse, D.J., Schweitzer, P.K., & Moul, D.E. (2005). Clinical pharmacology of other drugs used as hyponotics. In M.H. Kryger & W.C. Dement (Eds.), Principles and Practice of Sleep Medicine Philadelphia: Saunders. Carney, R.M., Freedland, K.E., Miller, G.E., & Jaffe, A.S. Depression as a risk factor for cardiac mortality and morbidity: A review of potential mechanisms. Journal of Psychosomatic Research, 53 597-902. Carpenter, J.S., & Andrykowski, M.A. (1998). Psychometric evaluation of the Pittsburgh Sleep Quality Index. Journal of Psychosomatic Research, 45, 5-13. Coates, T.J., Killen, J.D., George, J., Silverman, S., Marchini, E., & Thoresen, C.E. (1982). Estimating sleep parameters : A multitrait-multimethod analysis. Journal of Consulting and Clinical Psychology, 50 345-352. Cole, R.J., Kripke, D.F., Gruen, W., Mullaney, D.J., & Gillin, J.C. (1992). Automatic sleep/wake identification from wrist activity. Sleep, 15 461-469. 87

PAGE 88

Cross, N., Vazquez Sowell, L., Stutts, L., Smit h, K., Miles, W., Conti, J., et al. (2006). Secondary insomnia, depressive symptoms, and sleep self-efficacy in cardiovascular disease patients. Poster presented at the 20 th Anniversary Meeting of the Associated Professional Sleep Societies, Salt Lake City, UT. Currie, S.R., Wilson, K.G., Pontefract, A.J ., & deLaplante, L. (2000). Cognitivebehavioral treatment of insomnia secondary to chronic pain. Journal of Consulting and Clinical Psychology, 68 407-416. Dunbar, M., Ford, G., Hunt, K., & Der, G. (2000 ). A confirmatory factor analysis of the Hospital Anxiety and Depressi on scale: Comparing empiri cally and theoretically derived structures. British Journal of Clinical Psychology, 39, 79-94. Eaker, E.D., Sullivan, L.M., Kelly-Hayes, M., DAgostine, R.B., & Benjamin, E.J. (2005). Tension and anxiety and the prediction of th e 10-year incidence of coronary heart disease, at rial fibrillation, and tota l mortality: the Framingham Offspring Study. Psychosomatic Medicine, 67 692-696. Edell-Gustafsson, U., Svanborg, E., Swahn, E. A gender perspect ive on sleeplessness behavior, effects of sleep loss, and coping resources in patients with stable coronary artery disease. Heart and Lung, 35 75-89. Egede, L.E. (2007). Major depression in individuals with chronic me dical disorders: Prevalence, correlates and asso ciation with health resource utilization, lo st productivity and functional disability. General Hospital Psychiatry, 29 409 416. Gangwisch, J.E., Heymsfield, S.B., Boden-Albala, B., Buijs, R.M., Kreier, F., Pickering, T.G., et al. (2006). Short sleep durati on as a risk factor for hypertension: analyses of the first National Health and Nutrition Examination Survey. Hypertension, 47 833-839. Gerber, P.D., Barrett, J.E., Barrett, J.A., Ox man, T.E., Manheimer, E., Smith, R., et al. The relationship of presenting physical compla ints to depressive symptoms in primary care patients. Journal of General Internal Medicine, 7 170-173. Glassman, A.H., & Shapiro, P.A. (1998). Depr ession and the course of coronary artery disease. American Journal of Psychiatry, 155 4-11. Gomy, S.G., & Allen, R.P. (1999). What is an activity count? A comparison of different methodologies used in wrist actigraphy Poster presented at the 12 th Annual Meeting of the Associated Professional Sl eep Societies, Orlando, FL. Gorny, S.W., Allen, R.P., Krausman, D.T., & Cammarata, J. (2001). Initial demonstration of the accuracy and utility of an ambulatory, three-dimensional body position monitor with normals, sleepwalkers and restless legs patients. Sleep Medicine, 2 135-143. 88

PAGE 89

Guilleminault, C., Connolly, S.J., & Winkle, R.A. (1983). Cardiac arrhythmia and Conduction disturbances during sleep in 400 patie nts with sleep apnea syndrome. American Journal of Cardiology, 52 490 494. Haffajee, C.I., Chaudhry, M., Casavant, D., & Pacetti, P.E. (2002). Efficacy and tolerability of automatic nighttime atrial fibrillation shocks in patients with permanent internal atrial defibrillators. The American Journal of Cardiology, 89, 875878. Hegel, M.T., Griegel, L.E., Black, C., Goulde n, L., & Ozahowski, T. (1998). Anxiety and depression in patients receiving impla nted cardioverter-defibrillators: A longitudinal investigation. International Journal of Psychology in Medicine, 27 57 69. Herrmann, C., Muhen, F., Schaumann, A., Buss, U., Kemper, S., et al. (1997). Standardized assessment of psychological well-be ing and quality of life in patients with implanted defibrillators. Pacing and Clinical Electrophysiology, 20 95-103. Hlatky, M.A., Boineau, R.E., Higginbotham, M. B., Lee, K.L., Mark, D.B., & Califf, R.M. (1989). A brief self-adminis tered questionnaire to determine functional capacity (the D uke Activity Status Index). American Journal of Cardiology, 15, 651-654. Kanagala, R., Murali, N.S., Friedman, P.A ., Ammash, N.S., Gersh, B.J., Ballman, K.V., et al. (2003). Obstructive sleep apnea and the recurrenc e of atrial fibrillation. Circulation, 107, 2589-2594. Koch, C.G., Khandwala, F., Cywinski, J.B., Ishwaran, H., Estafanous, F.G., Loop, F.D., et al. (2004). Health-rel ated quality of life after coronary artery bypass grafting: A gender analysis using the Duke Activity Status Index. Journal of Thoracic and Cardiovascular Surgery, 128, 284-295. Konstam, V., Colburn, C., Butts, L. (1995). The impact of defibrillator discharges on psychological functioning of impl antable cardioverter defibrillator recipients. Journal of Clincal Psychol ogy in Medical Settings, 3 69-78. Kuhl, E.A., Dixit, N.K., Walker, R.L., Conti, J.B., & Sears, S.F. (2006). Measurement of patient fears about implantable cardi overter defibrillator shock: An initial evaluation of the Florida Shock Anxiety Scale. Pacing and Clinical Electrophysiology, 29 614-618. Kupfer, D.J., Reynolds, C.F., Ulrich, R.F., & Grochocinski, V.J. (1986). Comparison of automated REM and slow-wave sleep analysis in young and middle-aged depressed subjects. Biological Psychiatry, 21 189-200. Lacks, P., & Morin, C.M. (1992). Recent adva nces in the assessmen t and treatment of insomnia. Journal of Consulting and Clinical Psychology, 60 586-594. 89

PAGE 90

Lattimore, J.L., Celermajer, D.S., & Wilcox, I. (2003). Obstructive sleep apnea and cardiovascular disease. Journal of the American Co llege of Cardiology, 41 14291437. Lavery, C.E., Mittleman, M.A., Cohen, M.C., Muller, J.E., & Verrier, R.L. (1997). Nonuniform nighttime distribution of acute cardiac events: A possible effect of sleep states. Circulation, 96 3321-3327. Lavidor, M., Weller, A., & Babkoff, H. ( 2003). How sleep is related to fatigue. British Journal of Health Psychology, 8, 95-105. Lesperance, F., Frasure-Smith, N ., & Talajic, M. (1996). Major depression before and after myocardial infarction: It s nature and consequences. Psychosomatic Medicine, 58 99 110. Lesperance, F., Frasure-Smith, N., & Theroux, P. (2000). Depression and 1-year prognosis in unstable angina. Archives of Internal Medicine, 160 1354-1360. Lichstein, K.L., Wilson, N.W., Noe, S.L., Aguill ard, R.N., & Bellur, S.N. (1994). Daytime sleepiness in insomnia: Behavioral, biological, and subjective indices. Sleep, 17, 693 702. Lichstein, K.L., Wilson, N.W., & Johnson, C.T. (2000). Psychological treatment of secondary insomnia. Psychology and Aging, 15 232-240. Lichstein, K.L., Durrence, H.H., Taylor, D.J., Bush, A.J., & Riedel, B.W. (2003). Quantitative criteria for insomnia. Behavioral Research and Therapy, 41 427-445. Lichstein, K.L., Nau, S.D., McCrae, C.S., & Stone, K.C. (2005). Psychological and behavioral treatments for secondary insomnias. In M.H. Kryger & W.C. Dement (Eds.), Principles and Practic e of Sleep Medicine Philadelphia: Saunders. Lichstein, K.L., Stone, K.C., Donaldson, J., Nau, S.D., Soeffing, J.P., Murray, D., et al. (2006). Actigraphy validation with insomnia. Sleep, 29, 232-239. Lichtman, J.H., Bigger, J.T., Blumenthal, J.A ., Frasure-Smith, N., Kaufmann, P.G., et al. (2008). Depression and cor onary heart disease: Recommendations for screening, referral, and treatment. Circulation, 118, 1-8. Lisspers, J., Nygren, A., & Soderman, E. (1997). Hospital Anxiety and Depression Scale (HADS): Some psychometric data for a Swedish sample. Acta Psychiatrica Scandinavica, 96 281-286. McCrae, C.S., & Lichstein, K.L. (2001). Secondary insomnia: Diagnostic challenges and intervention opportunities. Sleep Medicine Reviews, 5, 47-61. 90

PAGE 91

McSweeney, J.C., Cody, M., OSullivan, P., Elberson, K., Moser, D.K., & Garvin, B.J. (2003). Womens early warning symptoms of acute myocardial infarction. Circulation, 108, 2619-2623. Means, M.K., Edinger, J.D., Glenn, D.M., & Fins, A.I. (2003). Accuracy of sleep perceptions among insomnia sufferers and normal sleepers. Sleep Medicine, 4 285-296. Mellinger, G.D., Balter, M.B., Uhlenhuth, E.H. (1985). Insomnia and its treatment: Prevalence and correlates. Archives of General Psychiatry, 42 225-232. Monti, J.M., & Monti, D. Sleep disturban ce in generalized anxiety disorder and its treatment. Sleep Medicine Reviews, 4 263-276. Morgan, K. (2000). Sleep and aging. In K.L. Lichstein & C.M. Morin (Eds.), Treatment of Late-Life Insomnia Thousand Oaks, CA: Sage Publications. National Heart, Lung, and Blood Institute Working Group on Insomnia. (1999). Insomnia: Assessment and management in primary care. American Family Physician, 59 3029 3039. National Institutes of Mental Health. Depression and Heart Disease. Retrieved on October 1, 2003 from the National Institutes of Mental Health Website: http://www.nimh.nih.gov/publicat/NIMHdepheart.pdf National Sleep Foundation. National Sleep Founda tion 2000 Omnibus Sleep in America Poll. Retrieved on February 1, 2005 from the National Sleep Foundation Website: http://www.sleepfoundation.org/ publications/2000poll.cfm Nishino, S., & Mignot, E. (2005). Wake-pro moting medications: Basic mechanisms and pharmacology. In M.H. Kryger & W.C. Dement (Eds.), Principles and Practice of Sleep Medicine Philadelphia: Saunders. Nowlin, J.B., Troyer, W.G., & Collins, W.S. (1965). The association of nocturnal angina pectoris with dreaming. Annals of Internal Medicine, 63, 1040 1046. Perlis, M.L., Jungquist, C., Smith, M.T., & Posner, D. (2005). Cognitive-Behavioral Treatment of Insomnia New York: Springer Publications. Piccirillo, J.F., Duntley, S., & Schotland, H. (2000). Obstructive sleep apnea. Journal of the American Medical Association, 284, 1492-1494. Pigott, T.A. (2002). Anxiety disorders. In S.G. Kornstein & A.H.Clayton (Eds.), Womens Mental Health: A Comprehensive Textbook New York: Guilford Press. 91

PAGE 92

Pycha, C., Calabrese, J., Gulledge, A., & Malo ney, J. (1990). Patients and spouse acceptance and adaptation to implantable cardioverter defibrillators. Cleveland Clinical Journal of Medicine, 57 441 444. Quesnel, C., Savard, J., Simard, S., Ivers, H., & Morin, C.M. (2003). Journal of Consulting and Clinical Psychology, 71, 189-200. Ranjan, P. (2005). Secondary insomnia. Retrieved on July 3, 2005 from the E-Medicine website: http://www.emedicine.com/med/topic3128.htm Redeker, N.S., Mason, D.J., Wykpisz, E., Glica, B., & Miner, C. (1994). First postoperative week activity patterns and recovery in women after coro nary artery bypass surgery. (1994). Nursing Research, 43, 168 173. Redeker, N.S., Mason, D.J., Wykpisz, E., & Gl ica, B. (1995). Womens patterns of activity over six months after coronary artery bypass surgery. Heart and Lung, 24, 502-511. Redeker, N.S., & Wykpisz, E. (1999). Eff ects of age on activity pa tterns after coronary artery bypass surgery. Heart and Lung, 28 5-14. Riedel, B.W., & Lichstein, K.L. (1998). Ob jective sleep measure and subjective sleep satisfaction: How do older adults with insomnia define a good nights sleep? Psychology and Aging, 13 159-163. Rogers, A.E., Caruso, C.C., & Aldrich, M.S. (1993). Reliability of sleep diaries for Assessment of sleep/wake patterns. Nursing Research, 42 368-371. Rybarczyk, B., Stepnaski, E., Fogg, L., Lopez, M., Paulette, B., Davis, A. (2005). A Placebo-controlled test of cognitive-behavioral therapy for comorbid insomnia in older adults. Journal of Consulting and Clinical Psychology, 73 1164-1174. Saunders, C. (2001). Hypertension: Controlling the Silent Killer. Boston: Harvard Health Publications. Schuster, P.M., et al. (1998). The psychosoc ial and physiological experiences of patients with an implantable cardioverter defibrillator. Rehabilitation Nursing, 23 30-37. Schmidt, N.B., Lerew, D.R., & Trakowski, J.H. (1997). Body vigilance in panic disorder: Evaluating attenti on to bodily perturbations. Journal of Consulting and Clinical Psychology, 65 241-220. Schweitzer, P.K. (2005). Drugs that disturb sleep and wakefuln ess. In M.H. Kryger & W.C. Dement (Eds.), Principles and Practi ce of Sleep Medicine Philadelphia: Saunders. 92

PAGE 93

Sears, S.F., Todaro, J.F., Saia, T.L., Sotile, W., & Conti, J.B. (1999). Examining the psychosocial impact of implan table cardioverter defibrillato rs: A literature review. Clinical Cardiology, 22 481-489. Serber, E.R., Sears, S.F., Sotile, R.O., Burns, J.L., Schwartzman, D.S., Hoyt, R.H., et al. (2003). Sleep quality among patients treated with implantable atrial defibrillation Therapy: Effect of nocturnal shock delivery and psychological distress. Journal Of Cardiovascular Electrophysiology, 14, 960-964. Sears, S.F., Shea, J.B., & Conti, J.B. (2005). How to respond to an implantable cardioverter-defibrillator shock. Circulation, 111, 380-382. Sears, S.F., & Conti, J.B. (2006). Psychologi cal aspects of cardiac devices and recalls in patients with implantable cardioverter defibrillators. American Journal of Cardiology, 15, 565-567. Spielman, A.J., & Glovinsky, P.B. (1991). The v.aried nature of insomnia. In P.J. Hauri (Ed.), Case Studies in Insomnia New York: Plenum. Spielman, A.J., Yang, C., & Glovinsky, P.B. (2005). Assessment techniques for insomnia. In M.H. Kryger & W.C. Dement (Eds.), Principles and Practice of Sleep Medicine. Philadelphia: Saunders. Spinhoven, P. Ormel, J., Sloekers, P.P., Kempen, G.I., Speckens, A.E., & Van Hemert, A.M. (1997). A validation study of the Hospital Anxiety and Depression Scale (HADS) in different groups of Dutch subjects. Psychological Medicine, 27 363370. Stepanski, E., Zorick, F., Roehrs, T., Young, D., & Roth, T. (1988). Daytime alertness in patients with chronic insomnia compared with asymptomatic control subjects. Sleep, 11 54 60 Stepanski, E.J., & Rybarczyk, B. (2006). Emerging research on the treatment and Etiology of secondary or comorbid insomnia. Sleep Medicine Reviews, 10 7-18. Suls, J., & Bunde, J. (2005). Anger, a nxiety, and depression as risk factors for cardiovascular disease: The problems and implications of overlapping affective dispositions. Psychololgical Bulletin, 131 260-300. Tabachnick, B.G. & Fidell, L.S. (2001). Using Multivariate Statistics (4 Ed.). Allyn and th Bacon: Boston. Taylor, D.J., Lichstein, K.L., & Durrence, H.H. (2003). Insomnia as a health risk factor. Behavioral Sleep Medicine 1, 227-247. 93

PAGE 94

Varkevisser, M., Van Dongen, H.P.A., & Kerkhof, G. A. (2005). Physiologic indexes in chronic insomnia during a constant routine: Evidence for general hyperarousal? Sleep, 12, 1588-1596. Varkevisser, M., & Kerkhof, G.A. (2005). Chronic insomnia and performance in a 24-hour constant routine study. Journal of Sleep Research, 14 49-59. Verbeek, I.H., Konings, G.M., Aldenkamp, A.P., Declerck, A.C., & Klip, E.C. (2006). Cognitive behavioral treatment in clinically referred chronic insomniacs: Group versus individual treatment. Behavioral Sleep Medicine, 4 135-151. Verrier, R.L., & Josephson, M.E. (2005). Cardiac arrhythmogenesis during sleep: Mechanisms, diagnosis, and therapy. In M.H. Kryger & W.C. Dement (Eds.), Principles and Practic e of Sleep Medicine Philadelphia: Saunders. Verrier, R.L., & Mittleman, M.A. (2005). Sleep-re lated cardiac risk. In M.H. Kryger & W.C. Dement (Eds.), Principles and Practi ce of Sleep Medicine Philadelphia: Saunders. Vorona, R.D., Winn, M.P., Babineau, T.W ., Eng, B.P., Feldman, H.R. et al. (2005). Overweight and obese patients in a primary care population repor t less sleep than patients with a normal body mass index. Archives of Internal Medicine, 165 25-30. Whang, W., Albert, C.M., Sears, S.F., Lampert, R., Conti, J.B., Wang, P.J., et al. (2005). Depression as a predictor for appropriate shocks among patients with implantable cardioverter-defibrillators: Results from the Triggers of Ventricu lar Arrhythmias (TOVA) study. Journal of the American College of Cardiology, 445 1090-1095. World Health Organization. (1990). International Statistical Classification of Diseases and Related Health Problems (ICD-10) (10 th ed.). Geneva: World Health Organization. Yashuda, Yoshida, Garcia-Garcia, Kay, & Krueger. (2004). Interleukin-1 has a role in cerebral cortical state-dependent electroencephalographi c slow-wave activity. Sleep, 28 117184. Zigmond, A.S., & Snaith, R.P. (1983). The Hospital Anxiety a nd Depression Scale. Acta Psychiatrica Scandinavica, 67 361-370. 94

PAGE 95

BIOGRAPHICAL SKETCH Natalie Joan Cross was born in 1977 in Albion, Michigan to her parents Patrick William and Cristine Marie Small and her older sister, Vanessa Cristine. She also has a younger sister, Audrey June. She was raised in Indiana and No rth Carolina. She graduated from D.H. Conley High School in Greenville, North Carolina in 199 5. Natalie earned a batchelors degree in psychology with a minor in Spanish in 2000 from No rth Carolina State University. She earned a masters degree in clinical psychology fr om East Carolina Univ ersity in 2004. Since 2004, Natalie has been a doctoral student in clinical and hea lth psychology at the University of Florida. She is currently a predoctoral psychology intern at the North Florida/South Georgia Veterans Administration Medical Center in Gainesville, Florida. Upon completion of internship, she will have fulfilled all requirements for her doctorate. Natalies career goals are to provide clinical and research services to patients with cardiac disease, sleep disorders, and post-traumatic stress disorder. 95