|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 EFFECTS OF EXERCISE ON FLUID BALA NCE INSTABILITY OF NYHA CLASS III AND IV HEART FAILURE PATIENTS By ANDREA MONIQUE BOYD 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 2008
2 2008 Andrea M. Boyd
3 To Brad, Aydan and Alexa, for enduring th e long nights, missed games, and absentmindedness. Thank you and I love you all dearly To my mother for whom without her unending encouragement and unconditional love I would never have made it to the front door. And, to my father, who taught me that learning was a cons tant and never ending challenge with boundless rewards.
4 ACKNOWLEDGMENTS I would like to thank my family for their continual support. I would like to thank my committee, Drs. James Jessup, David Criswell, Jennifer Elder, Eileen Handberg, and Laura Sutton, for providing such excellent guidance over th e past as this project moved and changed. I would also like to thank the Co llege of Medicine, Department of Cardiology, specifically Drs. Handberg, Pepine, and Schofield for their support of both my research proj ect and my education. I would like to thank the staff of GRECC at the VA, specifically Drs. Uphold, Carter, and Shorr for their support throughout thes e last trying months. And finally, I would like to thank Arlene Davis and Dr. Peggy Guin for their wonderful guida nce, support, and knowledge as I began the process that has now been fulfilled.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ............................................................................................................... 4 LIST OF TABLES ...........................................................................................................................7 LIST OF FIGURES .........................................................................................................................8 LIST OF ABBREVIATIONS ........................................................................................................ 10 ABSTRACT ...................................................................................................................... .............13 CHAPTER 1 INTRODUCTION .................................................................................................................. 15 Purpose ....................................................................................................................... ............15 Background and Significance .................................................................................................15 Hemodynamic Medical Management Model ..................................................................15 Measures of Hemodynamic Status ..................................................................................17 The Cardio-Renal Syndrome ........................................................................................... 20 2 LITERATURE REVIEW .......................................................................................................22 Physiological Framework ....................................................................................................... 22 Symptom Etiology of Heart Failure ................................................................................ 22 Fluid Balance Instability E tiology in Heart Failure ........................................................ 26 Exercise Training Effects in Heart Failure ............................................................................. 31 Summary ....................................................................................................................... ..........35 3 MATERIALS AND METHODS ...........................................................................................38 Original Study Florida Depa rtme nt of Health Study .......................................................... 38 Methods ..................................................................................................................................38 Design ........................................................................................................................ ......38 Specific Aims ..................................................................................................................38 Subjects ...................................................................................................................... ......39 Inclusion criteria .......................................................................................................39 Exclusion criteria ......................................................................................................39 Procedures .................................................................................................................... ...40 Variables ..................................................................................................................... .....43 Intervention .................................................................................................................. ....45 Dissertation Study ............................................................................................................ .......46 Methods ..................................................................................................................................46 Design ........................................................................................................................ ......46 Research Question ...........................................................................................................48
6 Subjects ...................................................................................................................... ......49 Procedures .................................................................................................................... ...51 Variables ..................................................................................................................... .....52 Data Analysis ...................................................................................................................53 4 ANALYSIS AND RESULTS ................................................................................................. 60 Demographics .................................................................................................................. .......60 Univariate Analysis: Sample ........................................................................................... 60 Univariate Analysis: Additional Characteristics ............................................................. 60 Specific Aim 1 ........................................................................................................................60 Univariate Analysis ......................................................................................................... 60 Bivariate Analysis ........................................................................................................... 62 Assumptions and Data Di stribution Analyses ................................................................. 63 Multivariate Analysis ...................................................................................................... 64 Specific Aim 2 ........................................................................................................................65 Descriptive Data .............................................................................................................. 65 Trend Analysis .................................................................................................................68 Variability Analysis ......................................................................................................... 72 Level of Change Analysis ...............................................................................................74 Summary ....................................................................................................................... ..........80 5 DISCUSSION .................................................................................................................... ...101 Discussion of Findings ........................................................................................................ .101 Summary of Data Analyses ...........................................................................................101 Sympathetic Nervous System, Fluid Balan ce Instability, and Exercise Training ......... 104 Endothelial Changes, Fluid Balance In stability, and Exercise Training ....................... 105 Summary ....................................................................................................................... .106 Limitations of the Study ...................................................................................................... .107 Methodological Limitations ..........................................................................................107 Statistical Limitations .................................................................................................... 108 Other Limitations ........................................................................................................... 108 Analyzing Data Using Statisti cal and Visual Analyses ........................................................ 110 Implications for Nursing Science ......................................................................................... 111 Implications for Clinical Practice ......................................................................................... 114 Recommendations for Future Research ................................................................................115 Physiological Model for Hemodynamic Fluc tuation Modulation in Heart Failure by Exercise Training ....................................................................................................... 115 Exercise Training and Endothe lial Dysfunction in Heart Failure related to Fluid Balance ....................................................................................................................... 115 Conclusions ...........................................................................................................................116 REFERENCES .................................................................................................................... ........118 BIOGRAPHICAL SKETCH .......................................................................................................135
7 LIST OF TABLES Table page 3-1 Baseline demographic and clinical characteristics ............................................................ 55 3-2 Cardiac medications upon discharge .................................................................................56 3-3 Summary of procedures .....................................................................................................58 3-4 Summary of Health Related Qual ity of Life (HRQOL) Measurements ............................ 59 4-1 Baseline characteristics : Admi ssion to discharge .............................................................. 82 4-2 Correlation summary table ................................................................................................ .83 4-3 Secondary variables correlation summary ......................................................................... 84 4-4 Summary of indivi dual characteristics ............................................................................... 85 4-5 Summary of subjects 1, 2 and 3 time variables ................................................................. 86 4-6 Summary of subjects 4, 5 and 6 time variables ................................................................. 86
8 LIST OF FIGURES Figure page 1-1 Heart failure: A cardiovascular-renal syndrome ................................................................ 21 2-1 Comprehensive, consensus physiologic model of heart failure sym ptoms and an overlay of areas that exerci se training has been shown to have a positive effect .............. 37 3-1 Study participants summary of events ...............................................................................57 4-1 Trend analysis for days exercised per week: Responders .................................................. 87 4-2 Trend analysis for days ex ercised per week: Non-responders ........................................... 87 4-3 Trend analysis for days exercised per week: Outliers ....................................................... 88 4-4 Trend analysis for amount of time sp ent exercising each session: Responders ................ 88 4-5 Trend analysis for amount of time spen t exercising each session: Non-responders .......... 89 4-6 Trend analysis for amount of time spent exercising each session: Outliers ...................... 89 4-7 Trend analysis for exercise inte nsity during each se ssion: Responders ............................ 90 4-8 Trend analysis for exercise intens ity during each session: Non-responders ...................... 90 4-9 Trend analysis for exercise inte nsity during each se ssion: Outliers .................................. 91 4-10 Variability analysis of daily weight: Responders .............................................................. 91 4-11 Variability analysis of daily weight: N on-responders .......................................................92 4-12 Variability analysis of daily weight: Outliers .................................................................... 92 4-13 Variability trend analysis for da ily we ight variability: Responders .................................. 93 4-14 Variability trend analysis for da ily we ight variability: Non-responders ........................... 93 4-15 Variability trend analysis for daily we ight variability: Outliers ........................................ 94 4-16 Level of change analysis for days exercised per week: Responders ................................. 95 4-17 Level of change analysis for da ys exercised per week: Non-responders ........................... 95 4-18 Level of change analysis for days exercised per week: Outliers ....................................... 96 4-19 Level of change analysis for amount of time exercise each session: Responders ............. 96
9 4-20 Level of change analysis for amount of time exercise each session: Non-responders ...... 97 4-21 Level of change analysis for amount of time exercise each session: Outliers ................... 97 4-22 Level of change analysis for exerci se intensity wh ile ex ercising: Responders .................98 4-23 Level of change analysis for exercise intensity while exerci sing: Non-responders .......... 98 4-24 Level of change analysis for exerci se intensity while exercising: Outliers .......................99 4-25 Level of change analysis for daily weight variab ility: Responders ................................... 99 4-26 Level of change analysis for da ily weight variabil ity: Non-responders .......................... 100 4-27 Level of change analysis for daily weight variab ility: Outliers ....................................... 100
10 LIST OF ABBREVIATIONS ACE-I Angiotensin converting enzym e inhibitor ADHERE Acute decompensated heart failure registry AICD Automatic implantable cardioverter defibrillator ANOVA Analysis of variance ANP Atrial natriuretic peptide ARB Aldosterone receptor blocker ARNP-BC Advanced registered nur se practitioner-Board certified AVP Arginine vasopressin BNP Brain natriuretic peptide BORG Borgs rate of perceived exertion scale BP Blood pressure BUN Blood urea nitrogen CABG Coronary artery bypass graft CO Cardiac output CP Chest pain DHEA Dehydrepiandostrone DOE Dyspnea on exertion EF Ejection fraction GH Growth hormone H2O Water HF Heart failure HFSA Heart failure society of America HR Heart rate HRQOL Health related quality of life
11 HRR Heart rate reserve IGF-I Insulin-like growth factor-I IR Insulin resistance LVAD Left ventricular assist device MI Myocardial infarction NE Nor-epinephrine NO Nitric oxide NS Not significant NYHA New York heart association PaCO2 Arterial carbon dioxide PAR Physical activities recall PCI Percutaneous coronary intervention PI Principal investigator QOL Quality of life RAAS Renin-angiotensi n-aldosterone-system ROS Radical oxygen species SA Sympathetic activation SD Standard deviation sICAM Soluble intercellular adhesion molecule SOB Shortness of breath SV Stroke volume TNF Tumor necrosis factor-alpha TX Transplant VCO2 Ventilated carbon dioxide VE Ventilation
12 VEGF Vascular endothe lial growth factor VO2 max Maximum ventilated oxygen VT Ventilatory threshold
13 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 EFFECTS OF EXERCISE ON FLUID BALA NCE INSTABILITY OF NYHA CLASS III AND IV HEART FAILURE PATIENTS By Andrea M. Boyd August, 2008 Chair: James Vernon Jessup Major: Nursing Science s The purpose of this study was to determine if a progressive, prescribed home-based aerobic exercise program would alter the natural physiol ogical processes that maintain fluid balance stability within the NYHA Class III/IV heart fail ure (HF) patient after medical optimization (titration of oral medical thera py with or without the infusion of an intrave nous inotrope). The study used a retrospective design applyi ng an innovative use of visual analysis typically reserved for single subject research. Subjects data fro m a previous study was utilized. These subjects were men and women over the ag e of 18 who had been diagnosed with HF via New York Heart Association (NYHA) for at least six months, and had been hospitalized for a current exacerbation in which they had been cl assified as NYHA III or higher. The exclusion criteria included receiving a CABG PCI, or having had an MI within the past six months prior to the hospital admission, having signif icant peripheral vascular dis ease, and/or being listed for heart transplant prior to the hospitalization (unless blood type O). The variables of interest for this study we re fluid status and me asures of exercise (intensity, duration, and frequency). The primary measurement variable for fluid status was daily body weight in pounds attained during the 24 weeks of the exercise intervention. Measures of actual exercise performed were grouped accordi ng to measures of intensity, duration, or
14 frequency. Exercise intensity was measured by rate of perceived exertion (BORG scale) while exercising. Exercise duration was measured via minutes of exerci se each exercise session. Finally, exercise frequency was measured by th e number of days exercised each week. To corroborate the exercise variables values, these variables were co rrelated with weekly pedometer readings (steps taken) for each week during the 24 week time frame for each subject. The intervention was a home based exercise pr otocol prescribed by the original principal investigator that lasted for 24 weeks. The exerci se participants had weekly phone calls to gather data and progress the exercise program and one 12 week follow-up. The usual care participants received random phone calls to collect exercise and pedometer data and had one 12 week followup visit to attain physical assessment values. A hierarchical multiple regression model significantly predicted daily weight fluctuations within a NYHA class III/IV heart failur e population after medi cal optimization (R2 linear= 0.713, F=3.224, p=0.015). The model determined that exercise intensity (BORG), exercise frequency (days exercised per week), and exerci se duration (minutes of exercise per session) directly predicted daily weight fluctuations (standard deviations of daily weights during a week time frame) when baseline weight fluctuati ons and event causing exit from the study were controlled. Overall, the model was predictive as theoretically indicate d, yet only 20% of the variance was accounted for. Th e retrospective visual analysis added depth to the model indicating that type of exercise and the needs of the patient should be accounted for when doing future research. In conclusion, this pilot study has demonstrated that exercise is a successful adjunctive therapy to managing the daily weight variability or fluid stat us instability of NYHA Class III/IV patients that is often a debilitating aspect of th e syndrome of heart failure.
15 CHAPTER 1 INTRODUCTION Purpose The purpose of this study was to determine if a progressive, prescribed home-based aerobic exercise program would alter the natura l physiological processes that maintain fluid balance stability w ithin the New York Heart Association (NYHA) Class III/IV h eart failure (HF) patient after medical optimization (titration of oral medical therapy with or without the infusion of an intravenous inotrope). Background and Significance The chronic illness of heart failure (HF) is a growing healthcare concern due to the rate of deaths attributed to HF increas ing 150% in 20 years (Goldfarb et al., 2004; Lowery, Massaro, & Yancy 2004). This is compounded by the fact that the United States population is growing older and the average age of onset or diagnosis of HF is 65 years of age (Lowery et al., 2004). Additionally, more than 90% of heart failure deaths occur in patients older than 65 years of age (Rich & Nease, 1999). With the high prevalence ra te and the associated high medical resource consumption, heart failure is now the single mo st costly cardiovascular illness in the United States (Rich & Nease, 1999). These economic factors, coupled with the increasing prevalence rate of HF, have driven research towards dete rmining methods to better manage heart failure both in the acute care setting and within the co mmunity, as well as, how to prevent the disease. Hemodynamic Medical Management Model In addition to heart failure (HF) being a chr onic illness that has a great economi c impact, HF is a syndrome that can affect anyone from birt h to the extreme elderly, of any race and either gender. Medical management of HF often require s a precision not readily or clearly discernable in the literature.
16 The most recent guidelines published by Heart Failure Society of America (2006), European Society of Cardiology (2005), and the joint guidelines by Amer ican Heart Association and the American College of Cardiology (2005), a ttempt to bring guidance to the heart failure practitioner using the most recent research and categorizing the evidence into the guidelines for best practice. These guidelines, although tiered based on type of patient being treated, have a few common aspects for all patients that bri ng together the common thought regarding the medical management of heart fail ure patients: the successful treat ment of heart failure patients relies on the balance of neurohormonal and he modynamic interventions (Abraham, 2007). The guidelines all stress the importance of asse ssing the patients hemodynamic status, which includes physical assessment (elevated jugular venous pressure, rales, the S3 gallop, hepatojugular reflux, edema, and as cites), the patients v ital signs (heart rate daily body weights, blood pressure (including orthosta tic), and temperature), the patient s symptoms (dyspnea at rest and during exertion, paroxysmal nocturnal dyspnea, orthopnea, fatigue, and nocturnal cough), a full serum panel (potassium, sodium, BUN, and cr eatinine), and depending on the status of the patient a variety of diagnostic test s (i.e. chest x-ray, echo, etc.). Over all, it is imperative to attain a full assessment of the patients hem odynamic and fluid volume status. Within the guidelines an overriding theme is to decongest and fluid optimize the patient while improving symptoms (often precipitated by congestion and fluid overload). With both the general management of the stable HF patient a nd the acute decompensate d HF patient attaining hemodynamic stability is the ove rall goal of the practitioner ("Executive summary: HFSA 2006 Comprehensive Heart Failure Practice Guideline", 2006). The t ools currently available are pharmacological interventions aimed at neurohorm onal blockades and diuretics aimed to flush the body directly of excess fluid once the patien t is in a state of fluid excess. These
17 interventions, although life saving, need improveme nts as the patients are becoming intolerant to the medications and polypharmacy has many draw backs of its own. Measures of Hemodynamic Status Although all guidelines currently recommend ph ysical assessm ent, vital signs, symptom assessment, and serum panels to measure a patient s hemodynamic status, there is little research to validate the use of some of these measures. Below is a list of the assessment techniques and a summary of the current research regarding the use of each technique as a measure of hemodynamic status within the HF population. Physical assessment The physical assessment techniques of rales, S3 gallop, elevated jugular venous pressure, hepatojugular reflux, ed ema, and ascites by the provider have recently received attention within the lite rature to determine if these techniques are useful in actually diagnosing HF. The ADHERE registry (Yancy Lopatin, Stevenson, De Marco, & Fonarow, 2006) reports that, depending on etiology of HF, between 67% and 69% of patients were admitted with rales and between 63% and 69% we re admitted with peripheral edema. This registry is the largest HF patient registry to date in the USA with over 187,000 patients registered as of 2006. Additionally, rales has been found to be 13% sensitive and 90% specific, the S3 gallop is 36% sensitive and 81% specific, JVP is 48% sensitive while 78% specific, edema is 10% sensitive and 94% specific, an d ascites is 15% sensitive and 92% specific all in diagnosing HF in the emergency setting (Yancy, 2007; Drazner, 2005). Overall, these techniques, in skilled hands, have been found to be helpful in the accurate diagnosis a nd treatment of acute decompensated HF (Yancy, 2007). Vital signs The measurement of heart rate, blood pressure (including orthostatic), temperature, and daily weights are often required every shift and at minimum every day while a patient is in the hospital setting. To date, heart rate and blood pressure have the most extensive
18 documentation. Yet, only heart rate and daily weight have demonstrated any correlation to acute decompensation events. There has been extensive research beginning in 1987 with Kannel which showed that a higher mean heart rate increased mortality from cardiovascular death. Research into physiologic variables and mortalit y continued, most recently with the remote monitoring available with implantable cardiac devi ces. The general findings are that both mean resting heart rate and mean heart rate increase as an acute decompensation event approaches approximately seven days prior to the hospitalizat ion. Additionally, such increases in heart rate increase all cause mortality, rehospitalization rates, and when such increases in heart rate occur at night both risks in mortality a nd rehospitalization increase at a greater rate (Ellery, Pakrashi, Paul, & Sack, 2006; Lechat et al., 2001; Reinlib & Abraham, 2003). As for daily body weight, research has shown mo st recently that an increase in daily weight values when taken at home by heart failure patients, was associated with hospitalizations for an acute exacerbation of heart failure (C haudhry, Wang, Concato, Gill, & Krumholz, 2007). Roth, et al (2004), found that ther e was a significant increase in th e trend for overall weight gain for thirty days prior to an admission for acute decompensation (r=.2) that remained at one week prior (r=.35) to the hospitalization. Additionally, those who were rehospitalized had a significant increase in both mean daily weight and mean h eart rate one week prior to rehospitalization (Roth et al., 2004). Finally, the majority of research in this area has been in the ar ea of managing patients using the transmission of these variable s. The overwhelming consensus has been that the transmission of heart rate, blood pressure and daily weights to a practitioner decreases hospitalizations, length of stay when hospitalized, hea lth resource utilizati on dollars, one year mortality, and anxiety, while increasing quality of life scores (Cleland, Louis, Rigby, Janssens, & Balk, 2005; Goldberg et al., 2003; Heidenreich, Ruggerio, & Massie, 1999; Roth et al., 2004).
19 Overall, heart rate, blood pressure, and daily bod y weight have been dire ctly linked to patient outcome if not directly linked to patient diagno sis, while body temperature has not been shown to have any relationship to heart failure events or the management of heart failure. Patients symptoms Patient symptoms have received attention as they have been the center of recent controversy. The controversy revolves around classification of dyspnea and the perception of dyspnea and therefore the subse quent alleviation of dyspnea. The ADHERE registry found that 34% of HF patients, regardle ss of etiology, were admitted with dyspnea at rest, whereas, 89% had general dyspnea (Fonaro w & Corday, 2004; Yancy, et. al., 2006). With such a large discrepancy between general dyspn ea and dyspnea at rest, discussion has begun to discern if dyspnea alone is a good in dicator of heart failure. To add to the controversy, dyspnea was found to have 50% sensitivity and 73% spec ificity (Yancy, 2007). Dyspnea during exertion, paroxysmal nocturnal dyspnea, orthopnea, fa tigue and nocturnal cough have no documentation as to whether these are valuable symptoms in the diagnosis of heart failure or not. Further documentation would be helpful to discern if thes e symptoms aid in the accurate diagnosis of HF. Serum panel. The four serum measures of potas sium, sodium, BUN, and creatinine are meant to measure serum fluid ba lance and how well the kidneys are able to maintain such balance. Using the ADHERE data, upon admission, HF patients had a mean sodium of 138 (SD not reported), a mean BUN of 34.5 (SD 22.5), and a mean creatinine of 1.9 (SD 1.8) (Yancy, et. al., 2006; Fonarow & Corday, 2004). Although the vari ability was large, both manuscripts noted that kidney function was impaired in many patient s or was at minimum in the beginning stages of becoming impaired. This finding has led the newest surge of re search into what is called the Cardio-Renal Syndrome in heart failure. The Cardio-Renal Syndr ome is the interplay
20 between the neurohormonal and hemodynamic intera ctions of the cardiovascular and renal systems that creates a negative f eedback loop leading to organ damage both in the kidneys and further damage in the heart. The Cardio-Renal Syndrome The cardio-renal syndrome is the triad of inte rwoven consequences of heart failure that result in vascular abnormalities, cardiac abnorm alities, and renal impairments (Francis, 2007; Konstam, 2007). The causes of each of these sets of abnormalities ar e lengthy. The vascular abnormalities ultimately result in decreased co mpliance and endothelial function. The cardiac abnormalities result in decreased compliance and c ontractility. Finally, the kidneys suffer from both the vascular and cardiac a bnormalities over time and begin to lose the ability to excrete sodium. This triad of abnormalities spirals upon itself and leads the patient down a path of hemodynamic instability. This often leaves th e practitioner playing th e balancing acting of polypharmacy and titration to attain the desired op tivolemic state, as well as, patient motivation towards desired lifestyle changes (i.e. dietary and activity). In tervening prior to the spiral downward and attaining a natura l hemodynamic homeostasis would add a powerful tool to the practitioners medical management arsenal. Exercise training has the potential to be that tool for this population due to its ability to affect th e neurohormonal system, the cardiovascular system, including the endothelial structure of the vasculature, a nd musculoskeletal system.
21 Figure 1-1. Heart failure: A car diovascular-renal syndrome NEURAL HORMONAL AND HEMODYNAMIC INSTABILITY CARDIAC ABNORMALITIES Decreased Compliance Decreased Contractility RENAL IMPAIRMENTS Decreased Sodium Excretion VASCULAR ABNORMALITIES Endothelial Dysfunction Decreased Compliance
22 CHAPTER 2 LITERATURE REVIEW Physiological Framework To fully understand the physiological as well as psychological and functional impacts that an exercise training program potentially has on the heart failure patient, a comprehensive understanding of the current physiological model of the symptoms of heart failure (HF) is warranted. Without this knowledge, it is difficult to comprehend the far reaching implications that a prescriptive exercise program can potentia lly have on the medical management of the HF patient. Symptom Etiology of Heart Failure The traditional theo ry of the etiology of th e symptoms of HF has been rewritten in the past 10 years. This change has occurred primar ily due to the innovative research heralding from the field of exercise science and research into d ynamic exercise and exercise training within the HF patient population and the HF animal model. Traditionally, the main symptoms of HF; dyspnea upon exertion (DOE), fatigue, exercise intole rance, and edema, were attributed to the failing heart and the vicious cycle of cardiac m yopathy, an increase in sympathetic activation, and a decrease in cardiac output (Bekheirnia & Schrier, 2006; Clark, 2006; Verbalis, 2006; Wong & Verbalis, 2002). With the extensive innovative and diverse literature that has emerged regarding circulating factors, muscular functi on, and respiratory func tion both during exercise and at rest within the HF population, a more refi ned and precise, yet not infallible, hypothesis of symptom etiology has emerged. Andrew Clark (2006) published his ideas regarding the origin of symptoms within HF. The hypothesis states skeletal muscle abnorma lities cause the majorit y, if not all of the symptoms demonstrated in HF (Clark, 2006).
23 The documented skeletal muscle abnormalities fa ll into one of two categories: structural alterations or metabolic alterati ons (Clark, 2006). Structural alterations include a shift from aerobic Type I fibers to more anaerobic Type II fibers (Drexler, Riede, Munzel, Konig, Funke, Just, 1992; Schaufelberger, Er iksson, Grimby, Held, & Swedbe rg, 1997; Sullivan, Green, & Cobb, 1990) and changes in the mitochondrial structur e of the skeletal muscle cell (Adams et al., 1999; Drexler et al., 1992; Dunnigan et al., 1987; Sullivan, et al, 1990). The metabolic alterations that have been documented are many. A decrease in overall oxidative capacity has been demonstrated (Drexler et al., 1992; Massie et al., 1988; Mettauer et al., 2001; Schaufelberger, et al, 1997; Sull ivan, et al, 1990) as well as a depletion in phos phocreatine and an increase in acidosis (Broqvist, Dahlstrom, Karlsson, & Larsson, 1992; Mancini et al., 1992; Massie et al., 1987; Massie et al., 1988). Additionally, the enzymatic activity has been demonstrated to be altered within the skeletal muscle cell (Drexler et al., 1992; Mancini et al., 1989; Schaufelberger, et al, 1997; Sullivan, et al, 1990). Neuronally, an increase in muscle cell sympathetic activity level has b een documented (Mitchell & Victor 1996; Roveda et al., 2003). Together, these skeletal muscle abnormalities, with others not yet discovered, cause a decrease in muscular bulk which in turn decreases muscular strength and endurance (Cicoira et al., 2001; Harrington et al., 1997; Minotti, Pillay, Oka, Wells Christoph, Massie, 1993; Volterrani et al., 1994) and are believed to be the limiting factor causing exercise in tolerance or muscular fatigue within the heart failure population (Adachi, Nguyen, Belardinelli, Hunter, Jung, Wasserman, 1997; Cicoira et al., 2001; Clar k, 2006; Corra, Mezzani, Giannuzzi, & Tavazzi, 2002; Harrington et al., 1997; Piepoli et al ., 2006; Piepoli, M.F., Ponikowski, Volterrani, Francis, & Coats, 1999; Vescovo et al., 1998).
24 Current researchers have suggested that the skeletal muscle abnormalities are a result of endothelial dysfunction (Nakamur a et al., 1995; Wilson, Ferraro, & Wiener, 1985) or vascular abnormalities. The endothelial dysfunction is pr oposed to be a result of the abnormally low bioavailability of nitric oxide (NO) within the endothelium of the HF pa tient (Adachi, et al., 1997; Agostoni & Bussotti, 2003; Maguire, Nugent, McGurk, Johnston, & Nicholls, 1998; Yoshida, Nakamura, Akatsu, Arakawa, & Hiramo ri, 1998) in combination with an increased formation of radical oxygen sp ecies (ROS) (Bauersachs & Scha fer, 2004; Drexler, 1999). Endothelial dysfunction results in four main dysfunctions in HF patients which include: 1) a thickening of the capillary basement membrane (Lindsay et al., 1994; Yoshida, et al, 1998); 2) a structural remodeling of the skeletal muscle vessels (N akamura, 1999), 3) a vascular hyporesponsiveness to stimulation (Indolfi et al ., 1994; Krum & Katz, 1998; Maguire, et al, 1998; Nakamura, Arakawa, Yoshida, Makita, Nii numa, Hiramori, 1998;); and 4) an impairment of the flow dependent dilatory response (A dachi et al., 1997; Kubo, Rector, Bank, Williams, & Heifetz, 1991; Nakamura, Yoshida, Arakawa, Mizunuma, Makita, Hiramori, 1996). It is this final dysfunction, an impairment of the flow de pendent dilatory response, which appears to affect the skeletal muscles most during exerti on. Blood flow does not increase adequately and the needed increase in oxygen does not reach the skeletal muscles subsequently causing apoptosis and the cellular abnor malities documented earlier (Adams et al., 1999; Vescovo et al., 2000). Endothelial dysfunction decreases the amount of general perfusion that the skeletal muscles receive causing skeletal muscle cell ab normalities (Nakamura et al., 1995). Continuing with Clarks (2006) hypothesis, he states that the abnormal sk eletal muscle cells cause an enhanced or overactivated ergoreflex response from the muscle receptors to exercise or work.
25 The ergoreflex is the reflex instigated by the ergoreceptors within the muscle which are sensitive to work performed (Piepoli, Adamopoulos, Bernar di, Sleight, & Coats, 1995). The stimulation of the ergoreceptors increases ventilation and sympathetic activation and it is this combined increase in ventilation and sympathetic activation that is called the ergoreflex (Piepoli, M. F., et al., 1999; Piepoli et al., 2006; Piepoli, M., Po nikowski, Clark, Banasiak, Capucci, & Coats, 1999). In HF this ergoreflex is more active (Pie poli, Clark, Volterrani, Adamopoulos, Sleight, Coats, 1996; Piepoli, M., et al., 1999). The over ac tivation of the ergoreflex increases ventilation (VE/VCO 2) or the ventilatory requirements to eliminate CO2 produced by metabolizing tissues (Piepoli et al., 1996; Piepoli, M.F ., 1999). The ventilation slope (VE on y axis by on VCO 2 x axis) is much steeper in HF patients as compared to healthy adults, and continues to become steeper with disease progression (Sullivan, Higginbotha m, & Cobb, 1988). Additionally, the ergoreflex drives the arterial carbon dioxide (PaCO2) too low. The bodys response in the HF syndrome is what is called dead space. Dead space is alveoli that are being ventilated but not perfused. Within the HF population, the increase in dead sp ace is thought to be a compensatory mechanism to counter the decrease in PaCO2 caused by the overactivated ergoreflex (Clark, 2006). Dead space has been shown to increase as the vent ilation slope increases supporting this concept (Clark, Volterrani, Swan, & Coats, 1997). This ov erstimulation of the ergor eflex, then initiates a cascade that, ultimately, is the downfall for th e HF patient symptomatically (Clark, 2006) and possibly neurohormonally. The activation of the ergoreflex also down re gulates the baroreflex (decreasing cardiac output and stroke volume) (Davies, Francis, Ju rak, Kara, Piepoli, & Coats, 1999; Piepoli 1996). This ultimately decreases the ejection fracti on which is a diagnostic criterion for the NYHA
26 staging system. Additionally, there is a decr eased vagal activation (A damopoulos et al., 1992; M. Piepoli et al., 1996) which yiel ds little to no counterba lance to the fourth and most significant trigger of the ergoreflexsympat hetic activation (Coats et al., 1992; Piepoli, et al., 1999). Fluid Balance Instability Etiology in Heart Failure Symp athetic activation has four major resu lting events: vasoconstriction of the renal vessels, hypothalamic stimulation an d release of arginine vasopre ssin (AVP), stimulation of the adrenal medulla and release of norepinephrine (NE), and stimul ation of glycogenesis coupled with insulin resistance (IR) (high plasma insulin levels with high fasting glucose levels) in previously non-IR individuals. All of these events have a cas cading neurohormonal response in the HF patient. Vasoconstriction of the renal vessels stimulat es the release of renin which stimulates release of Angiotensin II, often refereed to as the Renin-angiotensin-aldosterone system (RAAS) (Berl, Henrich, Erickson, & Schrier, 1979; Sc hrier & Abraham, 1999). Many pharmacologic interventions (i.e. ACE-inhibito rs) aim to stop the cascade at this point. Angiotensin IIs negative actions on the HF patient are vasoconstr iction of the periphery, which increases preload and increases wall stress with subsequent increa ses in left ventricular dilation and hypertrophy, and remodeling of cardiac myocytes, all which leads to or augments cardiomyopathy (Bekheirnia & Schrier, 2006; Schrier, 2006). Angiotensin II also promotes the reabsorption of sodium subsequently increasing plasma levels of sodi um (Selektor & Weber, 2007). Angiotensin II then stimulates the release of aldosterone (Bekheir nia & Schrier, 2006; Cadnapaphornchai, Gurevich, Weinberger, & Schrier, 2001). Aldosterone is a powerful neurohormone regulat or that triggers a cascade of effects. Aldosterone causes cardiac fibros is which contributes to cardiomyopathy via the remodeling of the cardiac myocytes (Schrier, 2006). Additionally, al dosterone increases sodium reabsorption,
27 and potassium and hydrogen ion excretion. Ultim ately water retention increases, increasing preload which increases wall stress, subsequently increasing left ventricular dilation and hypertrophy (Bekheirnia & Schrier, 2006; Cadna paphornchai, et al, 2001; Schrier, 2006). Aldosterone is also a sodium retaining hormone th at has been found to be elevated in edemetous HF patients yet when administered in high do ses to healthy adults does not induce edema (Hensen, Abraham, Durr, & Schrier, 1991; Wang et al., 2001). This finding underscores the importance that the physiology within the HF patient is not the same as the healthy adult. Finally, there is one other less understood effect of aldosterone and that is its relationship with growth hormone. This relationship will be furt her examined when discussing growth hormone resistance at a later point. Sympathetic activation also results in the stimulation of the hypothalamus and the subsequent release of arginine vasopressin (AVP) (Bekheirni a & Schrier, 2006; Schrier & Abraham, 1999). AVP is considered to be one of the most damaging circulating factors in HF (Bekheirnia & Schrier, 2006; Cadnapaphornchai, et al, 2001; Schrier, 2006). AVP causes an increase in the protein synthesi s of cardiac myocytes contributing to wall stress, cardiac myocyte remodeling, and the subsequent increases in left ventricular dilation a nd hypertrophy (Bekheirnia & Schrier, 2006; Schrier, 2006; Schrier & Ab raham, 1999). AVP also causes coronary constriction which leads to myocardial ischemia causing further wall st ress, left ventricular dilation and hypertrophy (Schrier, 2006; Schrie r & Abraham, 1999). AVP causes systemic arteriolar vasoconstriction increas ing after load thus increasing wall stress, left ventricular dilation and hypertrophy (Bekheirnia & Schrier, 2006; Cadnapaphornchai, et al, 2001; Schrier, 2006). Additionally, this arteriolar vasoconstriction also contributes to the activation of arterial receptors due to low cardiac output and further stimulates the earlier men tioned RAAS (Berl, et
28 al, 1979; Schrier, 2006). AVP also causes venoconstriction which increases preload, wall stress and subsequently increases left ventricular dilation and hypertrophy (S chrier, 2006; Schrier & Abraham, 1999). Finally, AVP stim ulates water retention, resulting in increases in preload, also increasing edema and the phenomenon of th ird spacing (Bekheirnia & Schrier, 2006; Cadnapaphornchai, et al, 2001; Schrier, Ohar a, Rogachev, Xu, & Knotek, 1998; Verbalis, Murase, Ecelbarger, Nielsen, & Knepper, 1998). Recent research into AVP receptor antagonists (i.e. tolvaptan) has demonstrated success in r ecent clinical trials (G heorghiade et al., 2004; Goldsmith, 2006) and shows promise as another pharmacologic interventional arena. The third event that sympathetic activatio n initiates is neurohormonal release of norepinephrine and stimulation of the adrenal medulla for serum release of norepinephrine (Clark, 2006). This causes an incr ease in resting heart ra te and, over the long te rm of the disease, decreases heart rate va riability (low peak HR with a high resting HR) (Kato et al., 1996; Keteyian et al., 1999; Larsen, Aukrus t, Aarsland, & Dickstein, 2004). The final event put into motion by the sympat hetic over activation is the increase in glycogenesis coupled with an increase in insuli n resistance which initiates a complex cascade of actions (Clark, 2006; Clark, Poole-Wilson, & Coats, 1996; Swan et al., 1997). This final event, increase in glycogenesis, increases the bodys meta bolic rate. This increase in metabolic rate subsequently initiates two further actions. First there is an increase in the catabolic steroid cortisol yet there is a decrease in dehydrepiandostrone (DHEA), the anabolic steroid precursor to cortisol (Anker, Chua et al., 1997). This rati o between cortisol and DHEA has been negatively correlated with body mass index demonstrating a catabolic-anabolic imbalance yielding muscle and fat loss (Anker, Clark et al., 1997). The second action is an increase in circulating growth hormone (GH) with a decrease in its effector hormone, insulin-like growth factor-I (IGF-I)
29 (Hambrecht et al., 2005; Hambrech t et al., 2002; Niebauer et al ., 1998). This imbalance between GH and IGF-I yields a GH resi stance (Anker et al., 2001). The high level of circulating GH enhances sympathetic activation. It is unknown as to why IGF-I is depressed while GH is elevated, ye t a relationship between aldosterone levels and growth hormone resistance has been suggested (Anand, Ferrari, Kalra, Wahl, Poole-Wilson, & Harris, 1989). The catabolic-anabol ic imbalance that results from the above mechanisms yields fatigue at the muscular level and exercise into lerance as a general symptom (Clark, 2006). The result is that it is the skeletal muscle that is the limiting factor not lung functi on, cardiac output or perfusion as previously thought, that causes the exercise in tolerance seen in HF patients (Cicoira et al., 2001; Clark, 2006 ; Corra, Mezzani, Giannuzzi, & Tavazzi, 2002; Harrington et al., 1997) All of the symptoms of HF have been acc ounted for within Clarks (2006) hypothesis. Others, though, hypothesize the cardiomyopathy seen in HF is a result of the Fragile Sarcolemma (Baudet, 2005). It has been suggest ed that the sarcolemma integrity of the cardiac myocyte has been compromised, resulting in th e remodeling and the larger pathology labeled cardiomyopathy (Lapidos, Kakkar, & McNall y, 2004; Toyo-Oka et al., 2004). Although not completely in sync with Clar ks hypothesis, it is very co mplementary. There are many contributing factors from Clarks hypothesis that could compro mise the integrity of the cardiac myocytes sarcolemma (i.e. e ndothelial dysfunction yielding d ecreased perfusion, increased protein synthesis, etc.). Some researchers would even say that the same mechanisms that alter skeletal muscle cells would also alter cardiac musc le cells. Additionally, it is noteworthy that in Clarks hypothesis there is no me ntion of the activation of the RAAS or of AVP release from sympathetic activation. This aspect is overlaid or drawn in from the research done on fluid
30 imbalance within the HF population. Although this research is not in contradiction to Clarks hypothesis, the leading researchers in the field of fluid balance re search attribute the sympathetic over activation to the decrease in cardiac output stimulating arteri al vessel receptors and not the ergoreflex activation (Cadnapaphornchai, et al., 2001; Schrier, 2006; Schrier, Cadnapaphornchai, & Umenishi, 2001). Overall, th ese hypotheses do not all agree as to the single trigger of the sympatheti c over activation. It is agreed, though, that the symptoms, not the disease process, are a direct result of this over activation (Baudet, 2005; Clark, 2006; Corra, et. al., 2002; Schrier & Fassett, 1998). Consequently, any intervention that could demonstrate a clear effect on this sympathetic over activatio n would therefore, hypot hetically, decrease the symptoms of HF. All the symptoms of HF have been incorpor ated within the above physiologic processes (dyspnea upon exertion, fatigue, exercise intolera nce, edema) and are visually demonstrated by the comprehensive, consensus model created by the author in Figure 2-1. There is one aspect of research that has no symptom or symptoms, yet has been heavily investigated in recent years pro-inflammatory markers. Within the HF patient, it has been demonstrated that tumor necrosis factor alpha (TNF ), interluekin-6 (IL-6), and soluble intracellular adhesion molecule-1 (sICAM) are elevated at rest (Cugno et al., 2005; Niebau er, Clark, Webb-Peploe, & Coats, 2005). It has been proposed that these are elevat ed due to endothelial damage wh ile others have suggested that these elevations are signs of the actual disease pr ocess itself (Fuchs & Dr exler, 2004; Gielen et al., 2005). A recent study demonstrated that it wa s not only an elevation of these markers but that there was also an impairment in respons e (Cugno et al., 2005). Cugnos (2005) study further supports the many other findings th at the pathophysiology of HF has some how altered how the body responds physiologically to internal factors, in this case, to the proinflammatory markers.
31 A side note in this area of research is that growth hormone treatment has shown a positive effect on these markers and is continuing to be pursued (Adamopoulos et al., 2002). Needless to say, this area of research needs further expansion and a greater depth of research to yield a better understanding of pro-inflammatory markers role within the physiology of HF and the possible role of pro-inflammatory markers in the development of symptoms. Exercise Training Effects in Heart Failure Early within the body of exercise science research and HF it wa s established that exercise training increased aerobic capacity, peak oxygen consumption, and the ability to tolerate exercise within the HF population. These findings have been summarized extensively due to the volume of studies that have been conducted (Smart & Marwick, 2004; van Tol, Huijsmans, Kroon, Schothorst, & Kwakkel, 2006). Research studies began expanding into the variety of plausible causes for such results. At the same time resear chers in the basic sciences began investigating animal models to explain the underlying mechanisms for the pathologic expression of symptoms demonstrated in HF. Using the proposed and graphically illustrated physiological model of HF symptoms in Figure 2-1, an understanding of the role that exer cise has as a treatment modality within the medical management of HF can be pursued. Following Clarks hypothesis of the over activation of the ergoreflex, exercise has demonstrated an effect on ventilation, down regulation of the baroreflex, and sympathetic over activation. Ventilation (VE/VCO 2). First, in regards to the increased ventilation dem onstrated in HF, exercise has been shown to directly decrease ve ntilation and decrease the slope of ventilation (VE/VCO 2) (Agostoni, Guazzi, Doria, & Marenzi, 2002; Armour, Clark, McCann, & Hillis, 1998; Giannuzzi, Temporelli, Corra, & Tavazzi, 2003; Gordon, Tyni-Lynne, Jansson, Kaijser,
32 Theodorsson-Norheim, & Sylven, 1997; Guazzi, Reina, Tumminello, & Guazzi, 2005; Witte & Clark, 2005). Additionally, it was research in to the mechanism behind dyspnea with exercise that led to the break through that it was not respiratory related dyspn ea but rather muscular fatigue that was the limiting event in exercise intolerance for HF patients. Baroreflex regulation Second, there is minimal research into the baroreflex up regulation seen with exercise. Exercise has demonstrated an increase in ejection fraction (Giannuzzi, et. al., 2003; Passino, et. al. 2006; van To l, et. al., 2006). At the same time exercise has shown an increase in left ventricle systol ic function (Belardinell i, Georgiou, Cianci, & Purcaro, 1999; Otsuka,et. al., 2003) and an incr ease in diastolic function (Meyer & LaederachHofmann, 2003). Although the studies are few, th ere is a trend demonstrating that exercise training does indeed up regulate the baroreflex in the HF population. Sympathetic nervous system. Finally, exercise has systematically demonstrated a decrease in sympathetic over activat ion via a variety of out comes. Heart rate variability has been improved by demonstrating a decrease in resting heart rate while peak heart rate has been increased (Cider, Tygesson, Hedberg, Seligman, Wennerblom, & Sunnerhagen, 1997; Keteyian et al., 1999; Larsen et al., 2004; Nishiyama et al., 2006). Resting levels of circulating norepinephrine levels have been decreased (Ketey ian et al., 1999; Kiilavuori, Naveri, Leinonen, & Harkonen, 1999; Passino et al., 2006; Tyni-L enne, Gordon, Jensen-Ursted, Dencker, Jansson, & Sylven, 1999), and resting levels of circulatin g aldosterone, angiotensin II, and AVP have all been decreased with exercise training (Brait h, Welsch, Feigenbaum, Kluess, & Pepine, 1999). Additionally, insulin resistance and hyperinsulie mia were both decreased (Nishiyama et al., 2006) and local expression at the skeletal muscle level of IGF-I was increased with exercise training (Hambrecht et al., 2005).
33 Skeletal muscle changes. Skeletal muscle abnormalities seen in HF patients have also been shown to be impacted with exercise training. Exercise training has increa sed the oxidative capacity of the skeletal myocyte (Gielen et al., 2005; Hambrecht et al., 1997), increased contractility response (Belardi nelli, 1998; Tyni-Lenne et al ., 1999), decreased the myocytes sympathetic neural activation (Rove da et al., 2003), and has instigated a shift from the anaerobic Type II fibers to the aerobic Type I fibers (Hambrecht et al., 1997). Additionally, exercise training increases bulk, strength, and endurance of the trained skelet al muscles (Minotti et al., 1993; Volterrani et al., 1994) and has improved the overall skeletal muscul ar function (Senden, Sabelis, Sonderland, Hulzebos, Bol, & Mosterd, 2005). And finally, when making the bridge between skeletal myocyte abnormalities and car diac myocyte remodeling, exercise has been shown to improve dysfunctional myocardium (B elardinelli, Zhang, Wasserman, Purcaro, & Agostoni, 1998). Endothelial changes In regards to the endothelial dys function demonstrated in the HF population, exercise training has demonstrated positiv e effects in various manners. Exercise has improved flow-dependent dilato ry response (Hambrecht et al., 1998; Kobayashi et al., 2003), increased blood flow to legs both during exercise and at rest (Hambrecht et al., 1998; Roveda et al., 2003), increased VEGF expression (Gusta fsson, Bodin, Sylven, Gordon, Tyni-Lenne, Jansson, 2001), and improved overa ll endothelial function (Hambrecht et al., 1998; Hornig, Maier, & Drexler, 1996; Kobayash i et al., 2003). And finally, although very minimal research has been conducted, it has been sh own that exercise increases th e basal availability of nitric oxide to the endothelium (Hambrecht et al., 1998). Other additional physiological factors. Exercise has demonstrated an effect on the physiology of HF in other areas. Exercise training has decreased circulating levels of TNF and
34 sICAM (Adamopoulos et al., 20 01; Conraads et al., 2002; Larsen, Aukrust, Aarsland, & Dickstein, 2001) as well as decreased local expr ession of these same fa ctors at the trained extremity (Gielen et al., 2003). The decrease in ci rculating levels may relate to the relationship between TNF and NO metabolism on the vascular endothelial and smooth muscle level (Kelly & Smith, 1997; Yoshizumi, M., Perrella, M.A., Bu rrnett, J.C., Lee, M.E., 1993). It has been suggested that the change in the local expression was an indi cator of the skeletal muscle alteration to a more antioxida tive state (Fuchs & Drexler, 2004). To further support this, exercise was shown to decrease circulating leve ls of hypoxanthine, a pro-oxidant substrate, in HF patients (Niebauer, Clark, We bb-Peploe, Boger, & Coats, 2005 ). Furthermore, exercise training decreases resting levels of both brain natriuretic peptid e (BNP) and atrial natriuretic peptide (ANP) (Braith, et. al., 1999; Passino et al., 2006). Psychological factors Last, but not least, exercise has also demonstrated an affect on the psychological outcomes of HF on the individual. A large base of research has been conducted and has demonstrated that exer cise improves the quality of lif e of the HF patient (Gary, 2006; Giannuzzi, et al, 2003; Johanss on, Dahlstrom, & Brostrom, 2006; Kodiath, Kelly, & Shively, 2005; Meyer & Laederach-Hofmann, 2003; Oka, DeMarco, & Haskell, 2005; Passino et al., 2006; Quittan, Sturm, Wiesinger, Pacher, & Fialka -Moser, 1999; Smith et al., 2005; Wielenga et al., 1998; Willenheimer et al., 2001). Researchers have begun to test the plausible reasons for this improvement. Self-efficacy has been shown to be improved with exercise (Gary, 2006; Oka, Demark, & Haskell, 2005), depression has decreased (Gary, 2006), and hospital admissions have decreased after exercise traini ng (Giannuzzi, et. al., 2003; Keteyi an et al., 1999) demonstrating an effect of exercise training on not only the physiological but psycho logical and economical impacts of HF.
35 Advanced Heart Failure (Class III/IV). Specific to the HF population being studied (NYHA Class III and IV), there is limited research available regarding how they respond to exercise training. The research has shown that, li ke Class II and III, the more severe HF patients respond to supervised exercise training with an increase in peak oxygen uptake and increase in total amount of time able to ex ercise (Friemark, Schechter, Schwamenthal, Tanne, Elmaleh, Shemesh, Motro, & Adler; 2007; Conraads, Vanderheyden, Paelin ck, Verstreken, Blankoff, Miljoen, Sutter, & Beckers, 2007; Erbs, Linke, Gielen, Fiehn, Walther, Yu, Adams, Schuler, & Hambrecht, 2003; Quittan, et al, 1999; Sturm, Quittan, Wiesinger, Stanek, Frey, & Pacher, 1999). Additionally, researchers have also demonstrated an increase in stroke volume in the more advanced stage (NYHA Class III vs. Class II ) patients with supervised exercise training (Erbs, et. al., 2003). Finally, supervised stre ngth training increased th e muscular strength and endurance of the severe HF pa tient (average ejection fraction of 19%) and had a subsequent decrease in fatigue and dyspnea with exercise (Beniaminovitz, Lang, LaManca, & Mancini, 2002). Of note, currently there is no published research with home-based exercise program and the severe HF population. Summary As demonstrated, exercise has the potential to affect HF at the phys iological cause of the symptoms that plague these patients for the entire ty of their disease. Research needs to further examine at what intensity, duration, frequency, in what environment, and for how long an exercise program needs to be used to be benefi cial physiologically. Specifically, in regards to fluid status, exercise ha s clearly demonstrated a potential for an intervention at the sympathetic activation step (as documented earlier), as well as, specifically decreasing resting levels of circulating angiotensin II and al dosterone (Braith, et al, 1999). Yet, no direct studies have been conducted examining the effect exercise traini ng has on the physiological and/or symptomatic
36 expression of fluid imbalance or fluid balance inst ability, peripherally or centrally, often demonstrated in the HF patient. The lack of st udies coupled with the supporting research of the reasonable expectation that exercise would positiv ely affect fluid balance stability, drives the desire to explore this rese arch area and supports the importance of the proposed study
37 Figure 2-1. Comprehensive, consensus physiologi c model of heart failure symptoms and an overlay of areas that exercise training has been shown to have a positive effect (pale blue areas) NO bioavailability circulating ET-1 ROS Endothelial Dysfunction Thickening of capillary basement membrane Structural remodeling of SM vessels Vascular hyporesponsiveness Im p airment of flow de p endent dilator y Skeletal Muscle Abnormalities Type II fibers, Type I fibers in Mitocondrial structure oxidative capacity phosphocreatine depletion muscle sympathetic neural activity Activation of Er g oreflex Vagal Activation Ventilation Regulation of baroreflex Sympathetic Activation DYSPNEA EF CO SV Aerobic capacity Vasoconstriction of renal vessels Hypothalamus stimulation NE Metabolic rate via glycogenesis and IR HR ? Arrhythmias Cortisol/ DHEA GH/ IGF-I GH resistance FATIGUE Wt Loss CACHEXIA AVP Protein synthesis of cardiac myocytes Coronary constriction Systemic arteriolar vasoconstriction Venoconstriction H 2 O retention CARDIOMYOPATHY EDEMA RAAS Vasoconstriction of periphery Remodeling of cardiac myocytes Cardiac fibrosis H2O retention Uncontrolled SA/ Augmented SA CardiacRemodelling
38 CHAPTER 3 MATERIALS AND METHODS Original Study Florida De partment of Health Study The original study was a longitudinal study condu cted through the Colle ge of Medicines Division of Cardiovascular Medicine under Eileen Handberg, AR NP-BC, PhD as the principal investigator and funded by a Florida Departme nt of Health grant. Methods Design The study was a prospective, randomized, contro lled parallel group design to assess the efficacy and safety of a progressive home based exercise program in patients with advanced heart failure who were being discharged from the hospital after medica l optimization. Medical optimization was defined as the titration of oral medical therapy with or without the infusion of an intravenous inotrope. Patient selection took place from a large tertia ry hospital setting which housed a heart failure clinic a nd heart transplant program. Specific Aims Research hypothesis 1: A progressive home exercise p rogram will result in significant improvements in functional capacity as measured by VO2max and NYHA at 12 and 24 weeks when compared to controls. Research hypothesis 2: A progressive home exercise program with structured interactions will result in significant improvements in psychological measures of quality of life when compared to controls at 12 and 24 weeks. Research hypothesis 3: A six month structured progressi ve home exercise program will result in changes in lifestyle so that exercise improvements will be maintained at one year when compared to controls. Research hypothesis 4: A progressive home exercise inte rvention will result in fewer hospitalizations for worsening heart failure dur ing the year of follow up when compared to controls.
39 Subjects Inclusion criteria Male or female age 18 or greater HF classified as NYHA Cla ss III or IV on admission Etiology of heart failure is idiopa thic or ischemic heart disease Ejection fraction of 35% or less within the previous 6 months Duration of heart failure 6 months or greater Hospitalized for worsening heart failure Receiving standard medical therapy (ACE Inhi bition, digoxin, diuretics, beta blockers, aldactone) for HF as clinically indicated Exclusion criteria CABG, PTCA or MI within the previous 6 months Significant peripheral vascular disease Concomitant illness that would interf ere with completion of the protocol Listed for transplant (unless blood type O) Participation in another research pr otocol that is ev aluating medications Recruitment. Patients admitted to the heart failure (HF) service of the hospital who met the inclusion criteria and had no exclusions were recruited for possible part icipation in the study. Subject participation was on a voluntary basis. Written informed consent was obtained from the subjects prior to participati on according the institutional guidelines established by the Health Science Center Institutional Review Board (IRB) at the University of Florida. Consent for patient participation was also obtained from the patients physician. Consecutive patients were approached for possible participation. Participan ts were not restricted on the basis of gender or race.
40 Protection of human subjects. The study was reviewed and approved by the University of Floridas IRB, received yearly renewals, and is current with IRB approval until 07/01/08. In accordance with IRB rules and HIPPA laws th e study protects the personal and health information of the subjects as well as the he alth and well being of th e participants. The only vulnerable population enrolled was that of wo men of childbearing age and there were no known risks to her participation and full disclosure was given prior to participation in the study. All risks and benefits to participa tion including not receiving treatmen t (intervention) was explained at the onset prior to c onsent and all participants were able to withdraw at any point without recourse. All data was securely stored onsit e at the College of Medicine, Department of Medicine under lock and key while all data stor ed on databases was de-identified prior to being placed within the database. A code key for database subject identification is kept under lock and key at the College of Medicine, Division of Cardiovascular Medicine in a separate file and will be destroyed at the completion of the study. Summary statistics. There were fifty-six (56) subject s consented with two (2) screen failures resulting in fifty-four ( 54) subjects randomized to one of two arms of the study. Below are two tables of baseline charac teristics of the subjects prior to randomization and there were no statistical differences between the two groups on any charac teristic (one way ANOVA, p 0.05). Table 3-1 is a summary of dem ographic characteristics while Ta ble 3-2 is a summary of the medications the subjects were taking upon discharge from the hospital (i.e., after medical optimization). Procedures Time point one. Upon admi ssion to the hospital and after consent had been obtained, each subjects baseline assessments were obtained which included a physical examination,
41 medical history, classificati on of NYHA status upon admission, an activity recall for the previous four weeks prior to admission using the Stanford 7-day Physical Activity Recall, and a full assessment of the subjects cu rrent quality of life using the battery outlin ed in Table 3-3. In addition, health resource utilization was assessed for the previous 12 months. Time point two. If subjects received intravenous inotropes, once the infusion was complete, and either prior to di scharge or within 5 days of di scharge, all subjects received a bicycle protocol maximum ventilation (VO2max) test. A maximum or peak oxygen consumption test is an objective measure of an individuals exercise capacity. The test is directly related to the integrated functional capacity of the cardiova scular, pulmonary, and skeletal systems of the individual being tested. Duri ng this test patients were familiarized with the BORG exertion scale, shortness of breath (SOB) scale and chest pain (CP) scale for use and recording at home. Subjects were then randomized by a com puter generated randomization schedule to intervention or control. Interv ention patients then received a 45 -minute instruction period on the initial progressive home walking program, standa rd lifestyle management instructions, which included tips on dietary habits, sp ecifically fluid management, and instructions on medications and self-monitoring activities. The exercise portion of the pr ogram was tailored to each individual based on the results of the exercise test. Patients were initially instructed to exercise to a BORG scale of 11 to 13 and/or 40% of h eart rate reserve (HRR). Patients were then instructed on how to take their pulse and how to record it. Patients were ambulated in the hall and educated on how to record data us ing the pedometer and daily activity log. The control group received iden tical instructions on the use of the pedometer in addition to the identical education regard ing standard lifestyle manageme nt instructions, which included tips on dietary habits, specifically fluid manageme nt, and instructions on medications and self-
42 monitoring activities. As per usua l and/or standard care, all subjects are rou tinely encouraged to be as active as possible. Th e controls, though, did not receive the individualized exercise prescription or the detailed instru ction period on the intervention prot ocol of progressive exercise based at home to their individual ability level. Subjects were th en officially placed in the 1st week of the study phase and medical therapy at discharge was recorded. Time point three. For twelve weeks following time poi nt two, all subjects recorded on activity diaries at least once a month that had b een provided to them. Usual care subjects had pedometer data collected once a month, which were selected at random by the study coordinator. Subjects in the intervention group were contacted weekly by a staff member to determine activity levels, collect pedometer data and to progres s the walking program on a weekly basis if appropriate. All subjects returned for a follo w up visit at 12 weeks for a physical exam, VO2max Quality of Life assessment via psychological test battery, activity recall for the previous four weeks, medical therapy assessment and health resource utilization assessment. Intervention subjects exercise prescription was evaluated an d altered as needed depending on results from their individual exams. Those who tolerate d the walking program were progressed to a stationary bicycle plus walking based intervention. Time point four. For twelve weeks following time point three, all subjects recorded exercise time (minutes), distance (miles), intens ity (BORG), daily weight (lbs), resting HR, and peak exercise HR on activity diaries that had b een provided to them at least once a month. Usual care subjects had pedometer data collected once a month, which were selected at random by the study coordinator. Subjects in the interventi on group were contacted weekly by a staff member to determine activity levels, collect pedometer data and to progress the bicycle and walking program or walking only program on a weekly basi s if appropriate. All subjects returned for a
43 follow up visit at 24 weeks for a physical exam, VO2max, Quality of Life assessment via psychological test battery, activ ity recall for the previous f our weeks, medical therapy assessment and health resource utilization assessme nt. Exercise bicycles were returned by the subjects who had progressed to the bicycle protocol. Time point five. Beginning at time point four, all subjects were under the usual care protocol and were encouraged to continue to be as active as possi ble. Calls were made monthly to assess progress and pedometer data was assessed monthly on a random basis for both groups. All subjects then returned fifty-two (52) week s from discharge to complete a physical exam, VO2max, Quality of Life assessment via psychological test battery, activity recall for the previous four weeks, medical therapy assessment and health resource utilization assessment. This visit completed the study and participants received a certificate of completion from the PI. Flow chart of study participant events. Figure 3-1 is a flow chart of the original study participants and the events that evolved over the course of the fifty-two (52) week study. Summary of procedures. In Table 3-3 the procedures for the original study are summarized by the time point in which each procedure was conducted. Variables The original study had two prim ary variables: functional cap acity and quality of life. Functional capacity. Functional capacity is an im portant diagnostic and prognostic assessment in patien ts with HF. In the strictes t physiologic sense it is a persons ability to perform aerobic work, and it has been defined by the maximal oxygen consumption (VO2max). In a more general sense functional capacity is the ab ility to carry out activi ties of daily living and has been assessed using more practical assessments such as the ability to carry groceries, tie shoes, and walk stairs. While the latter may be more practical, the AHA Advisory on the assessment of functional capacity in Clinical and Research Applications recommends that based
44 on the evidence, in patients with stab le, chronic heart failure, peak VO2 and ventilatory threshold (VT) (defined by the exercise level at which vent ilation begins to increase exponentially for a given increment in VO2) are highly reproducible and recommended for this population. The primary outcome measures for this study were tota l exercise time (total amount of time spent on spent on the exercise test) and peak VO2. It was anticipated that bo th groups would have some improvements at time point three (twelve weeks) in exercise tolerance due to the medical optimization, but that the progr essive exercise program woul d cause significantly greater improvements at 6 months (time point four). Secondary outcome measures would be the retention of improved exercise tolerance one-year post discharg e. Exercise testing was done using a bicycle due to the gait in stability of these severely i ll patients. Based on the known diurnal variability in exercise tolerance, exercise testing was conducted at the same time of day, and two hours after any meal. Functional capacity was also assessed using the New York Heart Association Classification (NYHA) of Heart Failure. While th is method is known to be imprecise due to its subjective nature, it served as a reference to the pr evious literature. It was anticipated that, due to the physiologic effects of th e up-titration of medical therapy, specifically ACE Inhibitors and the use of inotropic agen ts, all patients would increase activit y during the first twelve weeks of the study. Activity levels as a measure of functional capacity were measured in two ways. The Stanford 7-day Physical Activit y Recall (PAR) was utilized as a subjective measurement of activities of daily living. Assessment of day to day functioning may be more practically relevant than a patients ability to perform a sub maximal exercise test at a clinic visit. While the prognostic ability of the VO2max is extremely valuable, it is important to determine the day to day
45 responses to the intervention which a VO2max test can not realistically measure. The second measure taken was a pedometer reading to provid e a more objective measure of daily activity. There were limitations to this methodology, but gi ven the distance monitoring that results from using a home-based program, the data should gro ssly correlate with the more rigid assessments of functional capacity. Patients were called monthl y to assess activity using the pedometer for a 1 week period. Patients did not know which week s would be sampled, resulting in a fairer representation of th eir activity levels. Quality of life. Health-related quality of life (HRQOL) was a primary endpoint of interest because HRQOL reflects a patients abil ity to function in a variety of life domains including physical, social, and emotional domains The original study inco rporated a variety of quality of life assessment tool s to emphasize efficiency, depth, and comprehensiveness of measurement. Table 3-4 summarizes the t ools used, the psychological dimensions being indirectly measured, and the rationa le the original PI used for incorporating the given tool in this particular study. Overall the ba ttery of tests were evaluating both state and emotional traits important to the condition and management of heart failure. Intervention Exercise protocol. The intervention for the original study was a home ba sed, prescribed, progressive aerobic exercise progra m. Using the results from the exercise test and the physical exam, the PI (a cardiology specialist) pres cribed a progressive wa lking program for the individual tailored to the indivi duals capabilities and needs. An exercise prescription has three components: intensity, duration, and frequency. Intensity was gauged by the BORG scale and subjects were instructed to exer cise to an intensity of eleven to thirteen (11-13) (BORG ranges from 6 to 20) or to use 40% of their heart rate reserve (HRR). Th e HRR technique, although very reliable in the normal popul ation, is relatively unreliable within this population due to
46 cardiac medications being taken by the patients. Duration was variable although subjects were asked to exercise a minimum of fi ve to ten (5 -10) minutes each day. As subjects were able to tolerate exercise they were instructed to increas e their duration. Frequenc y was also variable, yet again subjects were asked to exercise a minimum of five (5) days each week. The type of exercise the subjec ts participated in was an aer obic based protocol. In the initial twelve (12) weeks of the intervention, the subjects were as ked to walk as their form of structured exercise. During the second twelve weeks, if the s ubjects had tolerated the walking program, the subjects received a stationary bicy cle and were asked to both walk and ride the stationary bicycle. The subjects were allowed to choose the amount of time spent on each type of exercise with the exercise prescription remain ing the same. The addition of the bicycle was to add variety and to allow and incr ease in exercise time for those not able to walk for long periods of time. The subjects received verbal encouragement via weekly te lephone calls from the study coordinator. During the entire intervention, subjects recorded their daily activity on a daily log which included heart rate, daily weight, SOB scal e, CP scale, BORG scale, minutes exercised, distance exercised, and a rating for the day (1-5). Dissertation Study The dissertation study was a secondary analysis of the original studys collected data. The dissertation study obtained a nd holds a separate IRB and is treated as a separate study for all data analyses. Methods Design The presented study used a retrosp ective design applying an innovative new application of visual analysis typically rese rved for single subject research. This app lication is introduced
47 here by the student as it has not previously be en used in the literature. Additionally, new terminology to characterize this approach has be en introduced to set it apart from traditional visual analysis and is termed re trospective visual analysis (RVA). Rationale: Retrospective visual analysis. A retrospective approach was the best approach due to the pilot nature of the research question and the availability of the detailed, indepth database from the original study. The purpose of using retros pective visual analysis, was to gain a greater depth of understa nding of how exercise and the degree of exercise may have influenced fluid balance stability in the severe HF patient beyond the statis tical or non-statistical significance of a statistical model. Traditional single subject methodology (SSM) is a quantitative, experimental methodology originally employed within the Behavi oral Analysis field of Psychology. SSM has only recently been used within the medical clinical research setting (Betker, Szturm, Moussavi, & Nett, 2006; Cohen, Chelland, Ball, & LeMura 2002; Effing, van Meeteren, van Asbeck, & Prevo, 2006; Feliz, Witt, & Harris, 2003; Fredri ksen & Mengshoel, 2000; Gross, Steinhauer, Zajac, & Weissler, 2006; Hammer et al., 2005; Huang, Chen, & Chung, 2006; Michaud & Nawoczenski, 2006; Morey, Cilo, Berry, & Cusick, 2003; Stoykov, Stojakovich, & Stevens, 2005). Traditional SSM visual analysis uses a syst ematic approach to plotting the data generated to guide the intervention timing a nd subsequent phases of the study. In doing so, a view of the data not typically observed in statistical analysis is gained with all data observed, no regression to the mean occurring, all variance observed, and all intervention effects demonstrated. The differences between traditional SSM visual analysis and retrospective visual analysis are: the data do not generate study phase changes,
48 the approach is used retrospectively ve rsus in real time or prospectively, statistical analysis may be used to generate subject selection for plot ting and profiling, and data analysis is used to further inve stigate or probe research question(s). The benefits to the new data analysis approach lie in the following: data plotted retains original expression dictating no regression to the mean, full variance expressed, and effects of interv ention or time demonstrated, the approach can be used retrospectively, a more in depth analysis of the research question may be obtained, the analysis is hypothesis genera ting for future research, and the analyses can be used in conjunction with st atistical analysis to provide a more complete understanding of the data available. The attributes of RVA are both virtuous for th e pilot researcher and a gold mine for the retrospective researcher as well. Both aspects made it key to the understanding of the data set this researcher was examining. The original study was a randomized, controlled group design, warranting a traditional statistical analysis in addition to RVA. With the employment of hierarchical multiple regressi on, a full predictive st atistical model was conducted without the loss of subjects or data. Togeth er statistical and visual analys is yielded the greatest amount of information available from this data set. The combination of both techniques provided a predictive statistical model while RVA yielde d in depth information regarding outliers, responders, and non-responders. Consequently, th e strongest foundation for understanding the data and the research question was produced. Research Question The proposed research question for this dissertation project was: o Does an exercise pr escription (intensity, duration, and frequency) predict fluid balance instability (daily weight vari ability) within the NYHA Class III/IV heart failure population after hospital based medical optimization?
49 Specific aims. The specific aims related to th e above research question were: I. To examine the relationship of the exer cise prescription variables (exercise intensity, duration, and frequency) on fl uid balance instability (daily weight variability) for all par ticipants during the study. II. To compare and contrast the trends, va riability, and level of change of the exercise prescription variables (exercise intensity, duration, and frequency) on fluid balance instability (daily wei ght variability) of the outliers, nonresponders, and responders according to the hierarchical multiple regression. Subjects The original study consented fifty-six (56) subjects at tim e point one (1), contributed fifty-four (54) subjects at tim e point two (2)--randomization, fo rty (42) subjects at time point three (3)twelve weeks, and thirty -one (31) at time point four (4)twenty four weeks to this dissertation study. The dissertati ons study has separate criteria for subject inclusion for each type of data analysis. Statistical analysis inclusion/exclusion criteria. All subjects with data from the original study were included in the statistical an alysis portion of data an alysis. The statistical analysis used was hierarchical multiple regression. This statistical analysis is able to tolerate missing data therefore all subjects with data at any time point within the 24 weeks of intervention contributed to the final analysis. Retrospective visual analysis inclusion/exclusion criteria. Only subjects who meet the following inclusion and exclusi on criteria were included in the retrospective visual analysis portion of the study. The inclusi on criteria are the following: o Subjects must be identified as an outlier (statistical or non-statis tical), non-responder, or responder by the hierarchical multiple regression model. o Subjects must have participated throughout the 24 weeks of inte rvention of the study o Subjects must have 75% of the daily logs completed. The exclusion criteria are:
50 o The subject has fewer than th ree (3) baseline weights o Or the subject failed to include daily wei ght on more than 30% of the daily logs. The estimated sample size needed for sufficient RVA is six (6) total subjects which represents 11% of the consented samp le and 19% of the 24 week sample. Ethical considerations. The dissertation study obtained a second IRB approval letter from the UF IRB-1 that is valid until October of 2008 when it w ill need renewal. This IRB is under the non-human category as the data was de -identified prior to an alysis. This study followed all guidelines and laws set forth by UF IRB, HIPPA, and the International Declaration of Helsinki for the conduct of research with hum ans. The utmost care was taken to protect the subjects health information ev en though all personal identifying in formation had been removed. The information was stored on a laptop that is password protected, as well as, uses fingerprint encoding to turn the computer on and to open all statistical files and programs. The original study held all informed consents in the original f iles. The study coordinators from the original study obtained th e consent, the original PI wa s not a medical provider of the subjects prior to the study and was only a prov ider during the study. Additionally, the subjects could withdrawal consent at any poin t during the study without repercussion. The four pillars of ethics: autonomy, beneficen ce, non-maleficence, and justice, were met in both studies (Stanley, 1998). Autonomy was met via proper consent and continued focus on the subjects proper medical care over the outcome of the study (i.e. inte ntion to treat). The dissertation study held this in check via respecting the integrity of th e original study and maintaining data security. Beneficence, to do good, was attained in the improvement in symptoms in the subjects in the original study. In the second study, the retrieval of information will lead to the enhancement of the symptoms and quality of life of HF patients yields beneficence to the HF population if not to the di rect participants. N on-maleficence, to do no
51 harm, was achieved as there were no adverse events related directly to the studies in either case. Finally, justice was met as there was no discrimi nation in the original study, and, consequently, in the second study, based on the participants age, race, gender, social economic status, or educational background. Clinical equipoise, or the uncertainty regard ing the advantages and/ or disadvantages of the intervention being tested, wa s not violated in either study (McCleary, 2002). Although an intervention was being tested and random assi gnment was conducted in the original study, both groups were educated on the basis of the in terventionphysical activity. The only difference was the manner in which the groups were followe d and treated regarding physical activity. The usual care group was not kept from exercising or even discourage d, as this was seen with many in this group performing their own home-based ex ercise programs. Again, as an intention to treat study, the usual care or control group received the best standa rd of care according to the current guidelines and, subsequently, clinical equipoise was not an issue for either study. Procedures Database. The original studys database was us ed for this dissertation study. The information for the variables of interest for this study were not of interest to the original study. The inform ation was captured and documented for medical and research purposes. However, the original database is qu ite extensive and retained a great deal of rich physiologic data pertaining to these subjects as they proceeded thr ough the fifty-two (52) weeks of the study Rationale for exclusion of time point five. The last twenty-eight (28) weeks of the original study were not included in the statistical analysis por tion of the dissertation study for two reasons. First and foremost, no interventi on was given during this timeframe. Second, the data collected was sporadic and inconsistent for th e variable of interest, fl uid balance variability. To adequately examine fluid shifts using daily weight measurements one must have consistent
52 daily weights over a week time period. W ithout such consistenc y, body weights are not accurately measuring fluid shifts. Theref ore, the final time point was excluded a priori to statistical data analysis. Variables There are two concepts being m easured in this study, fluid status and the exercise prescription. Fluid status is the primary or dependent variable and exercise prescription is the secondary variable or independent variable. Fluid status. The measurement variable for fluid status was daily body weight variability. Daily body weight wa s attained from four separate sources: hospital records during the initial hospital stay, self-report daily activity logs attained during the twenty-four (24) weeks of intervention, weights taken in termittently during the study at follow-up visits, and weights from clinical records. The daily weight variabil ity was assessed by taking the standard deviation of one weeks daily weights. The standard devi ation yields a variance of daily weight for the subject during a 7 day time frame indicating either a great amount of fluid shift or a static fluid shift. This technique was done to alleviate any increase or decrease in body composition that was attributable to exercise (i.e gain in muscle mass and/or loss of body fat) or lack of exercise (loss of muscle mass and/or gain or body fat). Body composition shifts are much slower and are not so easily detectable in variance within daily weight whereas, intracellular to extracellular fluid shifts are rapid and therefore more easily detectable with such mathematic techniques. Exercise prescription. The exercise prescription has as a basic formula: intensity, frequency, and duration. For the original study the pres cription had a fixed intensity (BORG of 11-13) with a variable session duration (minutes ) and variable frequency of days per week although that was fixed at as many as the patient could tolerate. With such a prescription, and the fact that the prescription was progressed as s ubjects tolerated exercise, the actual exercise
53 performed by each subject must be quantified to assess the effect of the prescription on fluid status. This was done so via a variety of ways. First, intensity was measured via the rate of perceived exertion (BORG scale) when exercising. Second, the dur ation of each exercise session was measured via the number of minutes of ex ercise each session. Third, the frequency was measured via the number of days exercised each week. Data Analysis Demographics. A univariate analysis was conducted at baseline to assess demographics characteristics. Specific Aim 1: To exa mine the relationship of the exercise prescription variables (exercise intensity, duration, and frequency) on fluid balance stability (daily weight variability) for all partic ipants during the study. A univariate analysis was conducted at baseline, twelve (12), and twenty four (24) week s to assess variables of interest. A bivariate analysis was conducted at baseline, twelve (12), and twenty four (24) weeks on all variables of interest to determine if any multicollinearity ex isted and to establish if any other important relationships existed prior to running the multivariate analysis. Assumptions for the multivariate analysis were tested prior to running the full model. Finally, a hierarchical multiple regression was run to determine if exercise intensity (a s measured by BORG), exercise frequency (as measured by days exercised per week), and ex ercise duration (as measured by minutes of exercise per session) directly predicted daily weight fluctua tions (as measured by standard deviations of daily weights duri ng a week time frame) when base line weight fluctuations and event causing exit from the study we re controlled with in the model. Specific Aim 2: To compare and contrast th e trend, variability, and level of change of the exercise prescription variables (exercise intensity, duration, a nd frequency) on fluid balance stability (daily weight variability) of the outliers, non-responders, and responders
54 according to the hierarchical multiple regression. A descriptive data summary was conducted for each subject. Additionally, a retrospective visual analys is was conducted for individual trend, variability, and level of change for each variable of interest for outliers, non-responders, and responders as determined by the regression model. Subjects were selected using a priori inclusion/exclusion criteria.
55 Table 3-1. Baseline demographic and clinical characteristics All Subjects (n=54) Usual Care (n=27) Home Based Exercise (n=27) Age (years) Mean (SD) 57.8 (11.8) 59.1 (12.4) 56.6 (11.4) Gender Male 42 (77%) 21 21 Race White 45 (83%) 27 18 BMI Mean (SD) 29.8 (5.9) 29.8 (5.4) 29.8 (6.4) Ejection Fraction (%) Mean (SD) 17.7 (7.0) 17.6 (7.6) 17.8 (6.5) HF Duration (months) Mean (SD) 55.6 (35.3) 47.4 (32.6) 64.5 (36.5) HF Etiology Ischemic 28 (52%) 13 15 Idiopathic 26 (48% 13 13 AICD 24 (44%) 8 16 Revascularization 24 (44%) 10 14 MI 22 (41%) 9 13 Atrial Fib 14 (26%) 7 7 Diabetes 24 (44%) 13 11 Hypercholesteremia 21 (39%) 9 11 Hypertension 29 (54%) 13 16 Smoking Status Past/Current Smoker 41 (76%) 19 22
56 Table 3-2. Cardiac medications upon discharge All Subjects (n=54) Usual Care (n=27) Home Based Exercise (n=27) Ace-inhibitors 40 (74%) 18 22 Beta-Blockers 48 (89%) 22 26 ARBs 15 (28%) 8 7 Diuretics 50 (92%) 24 26 Digoxin 42 (78%) 19 23 Aspirin 22 (41%) 10 12 Statin 31 (57%) 15 16 Nitrates 10 (18%) 7 3
57 Figure 3-1. Study participants summary of events Home Based Exercise n= 27 Usual Care n=27 n=21 n=19 Screen Failure n=2 2 deaths 1 heart tx 1 severity of illness 1 relocation 1 non-compliant 2 rehab 2 deaths 1 heart tx 1 relocation 1 non-compliant 1 unknown 2 deaths 1 heart tx 1 LVAD 1 other study 2 unknown 1 death 1 heart tx 1 LVAD 1 severity of illness 2 rehab 2 severity of illness 1 unknown 1 death 2 heart tx 2 severity of illness 1 unknown n=13 n=14 Time Point 5 Weeks (n=11) N=18 (n=7) Consented upon admission N =56 Time Point 2-Discharge Randomization N =54 Ti m e Point 3 12 Weeks Time Point 4 24 Weeks
58Table 3-3. Summary of procedures Procedures Entry Admission (Time Point 1) Discharge (Time Point 2) Week 12 (Time Point 3) Weeks 1-23 Week 24 (Time Point 4) Weeks 2451 Week 52 (Time Point 5) H&P X X Physical Exam X X X X X Informed Consent X Medical Optimization X X Exercise VO 2 max X X X X Quality of Life X X X X Randomization X Activity Assessments X X X X X X Health Resource Utilization X X X X Training Calls* X* X** *Weeks 1-23 exercise participants receive d weekly calls from study RNs about their progress and their exercise perscription **Weeks 24-51 all participants receive d random calls from study RNs abou t their progress and activity level.
59Table 3-4. Summary of Health Related Quality of Life (HRQOL) Measurements Quality of Life Dimension Rationale Instruments Physical Functioning Reflects ejection fraction indirectly SF-36 Physical Functioning1 SF-36 Physical Role Functioning1 Emotional Functioning Impact of surgery and medications Im pact of mental health history Impact of chronic morbidity SF-36 Emotional Health1 SF-36 Emotional Role Functioning1 SF-36 Vitality1 CES-D depression (20 items)2 STAI trait anxiety (20 items)3 Social Functioning Demonstrates ability to return to previous social relationships SF-36 Social Functioning1 Study questionnaire Symptom Reduction and Satisfaction Examines CHF-specific functioning and rela ted pain LVD-36 Total Score4 Minnesota Living with Heart Failure Questionnaire5 SF-36 Bodily Pain Scale1 Overall Well-Being and Health Summary measure Provides general com parison with specific dimensions SF-36 General Health1 Satisfaction with Life Scale6 Life Orientation Test optimism7 1McHorney, Ware, & Raczek, 1993; 2Weissman, Sholomskas, Pottenger, Prusoff, & Locke, 1977; 3Ramanaiah, Franzen, & Schill, 1983; 4O'Leary & Jones, 2000; 5Rector, Kubo, & Cohn, 1993; 6Diener, Emmons, Larsen, & Griffin, 1985; 7Scheier et al., 1989
60 CHAPTER 4 ANALYSIS AND RESULTS Demographics Univariate Analysis: Sample The total sample had 54 subjects and the ba seline dem ographics are summarized in the original study section (Chapter 3) of this di ssertation in Tables 3-1 and 3-2. The original subjects were selected from a convenience sample and were randomly placed into the exercise intervention or usual care group. For the following dissertation analysis, all of these subjects data were part of a secondar y, retrospective data analysis. Univariate Analysis: Addi tional Ch aracteristics In Table 4-1 there is a summary of baseline characteristics not previously summarized that are pertinent to this disse rtation study. For this sample the presence of JVD, peripheral edema, murmur, and cardiac S3 significantl y dropped from admission to discharge (68% to 34%) and rales and enlarged liver dropped mode rately during the same time frame (19% and 13% respectively). Specific Aim 1 To examine the relation ship of the exerci se prescription variables (exercise intensity, duration, and frequency) on fluid balance stability (daily weight variability) for all participants during the study. Univariate Analysis Dependent variable. Fluid status instabilit y was measured via daily weight variability. Each participants daily weights were recorded for one week time periods and then the standard deviation for each week was calculated. This numb er was recorded as the weekly daily weight variability for each subject during the 24 weeks of the intervention portion of the original study.
61 The grand mean for this vari able for all subjects was + /1.1 pounds (SD 0.95) with a range of 0 to 4.9 pounds meaning that most subjects varied + /1.1 pounds around their mean for a given weeks set of daily weight values for that 7 day period. Additionally, subjects ranged from having no variance to having a variance of + /4.9 pounds around their mean for that weeks set of daily weights. Control variables. All subjects exit event or the even t that caused them to leave the study were categorized into one of six categorie s: Completion of study (n=22), Non-compliant to protocol (n=2), Severity of illness (n=6), Received heart transplant or LVAD (n=8), Death (n=8), or Other (n=8). Within the model these variables were list ed in order of severity of exit with death being the last categ ory (6) and completion of study being the first category (1). The subjects initial diuresis while in the hospital for medical optimization was used as the baseline daily weight variability. This was calculated via taking all weights recorded while the subject was in the hospital and then calculati ng the standard deviation of those values. The grand mean for the baseline daily we ight variability was +/-6.4 pounds (SD 5.9) or 6.4 pounds around the individuals mean weight dur ing hospitalization with a range of + /1 to + /27.9 pounds around the individuals mean during hospitalization. Independent variables. Exercise intensity was measured via the BORG scale per exercise session. All BORG values for each week were averaged to yield a single BORG value for each week of the study. The BORG scale r uns from 6-20 with 6 representing no exertion and 20 representing the need to stop ex ercising. The grand mean was 11.8 (+ /1.8) with a range of 6.4-16.5. On the BORG scale 11 is correlated with the word Light whereas 13 is Somewhat Hard. The frequency of exercise was measured via the amount of days exercised each week. The grand mean was 3.4 days (SD 1.6) with a range of 0-7 days/week meaning that most subjects
62 exercised an average of 3.4 days a week with su bjects ranging from no exercise to exercising 7 days per week. Finally, the duration of exerci se was measured by the amount of time spent exercising each session which was measured in minutes. The amount of time for each session was averaged for each week of the study to determine the weekly time spent exercising each session value. The gran d mean was 13.2 minutes (SD 9.4) with a range of 0-39.5 minutes stating that the average exercise session was 13.2 minutes long with the range being no exercise to 39.5 minutes. Bivariate Analysis Primary variables. A co mplete correlational analysis including both biva riate and partial correlations (controlling for base line daily weight variability and exit events) was conducted. Relationships of significance are summarized in Table 4.2. Overall, when controlling for baseline variability in daily weight and exit events, daily weight variability was significantly correlated to the amount of days exercised such that as the amount of days exercised per week increased, the variability in daily weight decreased (r = 0.404, p=0.037). Although not statistically significant, a similar, moderate rela tionship held true with daily weight variability and the amount of time exercised each session ( r = 0.289, p=NS). To a lesser degree the intensity at which the participants exercised also increase d as the variability in daily weight decreased ( r = 0.156, p=NS). Also, of note is the inverse rela tionship between intensity and days exercised ( r = 0.220, p=NS), as the amount of days increased, th e intensity exercised decreased. This same relationship held true with time and intensit y, although when corrected for in the partial correlation the relation decreased ( r = 0.224 and r = 0.195, p=NS). Secondary variables. Bivariate correlations including th e primary set of variables and the 7-day Physical Activity Recall (PAR) and pedo meter readings and partial correlations were
63 run with exit event and daily weight variability at baseline being controlled. These bivariate analyses were conducted to dete rmine if there were any relationships between these objective and subjective measures and the meas ures of the exerci se prescription. The pedometer readings had di rect statistically significant relationships with the amount of time spent exercising each se ssion and the amount of days exercised each week in both bivariate and partial co rrelation analyses (Table 4-3). As the subjects increased the amount of time exercised and the amount of days exer cised per week the amount of steps taken each week increased. Exercise intensity did not dem onstrate a statistically significant relationship with pedometer readings. None of the PAR relationships were statisti cally significant.(Table 4-3). The pedometer by PAR relationship was not statistically signifi cant and had a small correlation both in the bivariate and partial analyses ( r =0.152 and r = 0.131, p=NS). Assumptions and Data Distribution Analyses Assumptions. The normality of distri bution of the error term assumption for the multiple regression was met as the error term met the Gau ssian normality of distribution criteria and did not require any transformation. The assumptions of linearity and homoscedeacity were both screened for and met (i.e. not violated). The a ssumption of independence of the error term was screened for and met with a Durbin-Watson of 2.05. Data distributions. There were no statistically significant outliers or influential cases. Yet, it is of note that this population is in a nd of itself a population of outliers from the normal heart failure population due to the state of illn ess the participants were in upon entering. Multicollinearity was not an issu e, no VIF was greater than 10.
64 Multivariate Analysis A hierarchic al multiple regression was run to determine if exercise intensity (as measured by BORG), exercise frequency (as measured by da ys exercised per week), and exercise duration (as measured by minutes of exercise per session) directly predicted daily weight fluctuations (as measured by standard deviations of daily we ights during a week time frame) when baseline weight fluctuations and event cau sing exit from the study were c ontrolled within the model. The model dictated that exercise intensity, fre quency, and duration, while c ontrolling for event of study exit and baseline daily weight fluctuation, significantly predicted daily weight fluctuations within a NYHA class III/IV heart failur e population after medi cal optimization (R2 linear= 0.713, F=3.224, p=0.015). The model accounted for 20% of the variance se en in daily weight fl uctuations (adjusted R2=0.2). Within the model, the amount of days ex ercised accounted for the greatest effect on the variance within the model which was statisti cally significant (=-0.608, p=.008). As the amount of days exercised per week increased, the variability in daily weight decreased. The amount of time exercised had a similar effect that was al so statistically signifi cant (=-0.426, p=.019). As the amount time exercised per session increased, the variability in daily weight decreased. Exercise intensity as measured by BORG had minimal effect and was not a statistically significant predictor of the vari ance (=-0.138, p=NS). As BORG in creased (or as intensity of exercise increased), daily weight variability decreased although this effect was not statistically significant. The control variables had a mixed effect. When controlling for the exit event, there was a minimal effect and it was not statistically significant (=-0.208, p=NS). The events were coded such that as the events became larger it indicated more severe exit events. Therefore, as the exits events became more negative (i.e. tran splantation or death), daily weight variability increased. This was again not statistically significant as a predictor of variance within the model.
65 The baseline daily weight variability (hospitali zation diuresis during medical optimization) was a significant predictor of varian ce within the model (=0.469, p=.003). As the baseline daily weight variability increased so did th e weekly daily weight variability. Specific Aim 2 To compare and contrast the trend, variabil ity, and level of cha nge of the exercise prescription variables (exercise intensity, duration, and frequenc y) on fluid balance stability (daily weigh t variability) of outliers, non -responders, and responders according to the hierarchical multiple regression. There were six (6) subjects whom met the incl usion criteria for the retrospective visual analysis. RVA was conducted on all six subjects There were two outliers, one true and one explained by circumstance. There was one tr ue non-responder and one non-responder according to the regression model, but when graphed out to completion, this subject was actually a late responder. Finally, there were two true responders. Each subject will first have descriptive data presented and then will be graphed with textual commentary to follow each graph. Descriptive Data Subjects one (1) and two (2) were considered true responders by the regression model. This m eant that on all three vari ables (exercise intensity, duration, and frequency) these subjects had a decrease in daily weight variability such that on a three dimensional model these values fell very close to the proposed regression line. Subjects three (3) and four (4) were considered non-responders by the regression. This means that these subjects did not respond as expected according to the proposed regression model. Subject three, once graphed through completion, turned out to be a late responder. The regression model only takes in data up to twenty four weeks. RVA allows for all data to be assessed. Subject three had a re sponder trend, just after twenty four weeks, therefore is termed a late-responder. Finall y, subjects five (5) and six (6)
66 were deemed outliers (although not statistically) to the regression model. Using Mahalanobis values, those that were the farthest from the mode l were chosen, resulting in subjects five and six who met all inclusion and exclusion criteria. Subject five provided a great deal of valuable information. Although numerically an outlier, when graphed to completion this subject no longer acted like an outlier. Subject six is a true ou tlier in multiple aspects and displayed this both graphically and textually. To understand the subjects more clearly, a su mmary of baseline char acteristics has been presented in Table 4-4. To note, all of these subjects were male and white although the sample was not entirely male or all whit e. In addition to these summary characteristics, none of these subjects were working. There are also some important individual char acteristics to know a bout these subjects. Subject one was placed in the hospi tal at week thirteen (13) to await a heart transplant. He received this transplant at week twenty six ( 26). Subject two (2) was a veteran who received care through the VA as well as the heart failure clinic. It is no t known if he was a part of the TeleHeart program while particip ating in this study, which doe s telephone follow up for weight management as well as other HF symptom mana gement. Subject three (3) was an exercise subject who was transitioned to the bicycle prot ocol at week twelve ( 12). He returned his bicycle at week twenty five (25). His initia l walking exercise prescription was 5-10 minutes a day, 5 days/week, at 11-13 BORG or 106 HR. He wa s increased at twelve weeks to bicycling 30 minutes a day, 3-4 days/week, at 11-13 BORG or 105 HR and walking, in addition to bicycling (if tolerated), 15 minutes a day, 1-2 days a week, at 8-10 BORG. From his exercise logs, during the second twelve weeks when he documented which type of exercise he was doing, he was doing bicycling. Unfortunately, he did not always document the type of exercise through-out.
67 Subject four (4) was a subject who immedi ately placed in a cardiac rehabilitation program. He participated in this cardiac reha bilitation program from his discharge until week seventeen (17). Subject five (5 ) was an exercise subject who al so increased his protocol from walking to bicycling at twelve weeks. At what point he returned his bicycle is unclear, although the intervention portion of the study was officially complete at w eek twenty four (24). His initial exercise prescription was walki ng 15 minutes a day, 5-6 days/wee k, at a BORG of 11-13 or HR of 92. At twelve weeks he was prescribed bicy cling 45 minutes a day, 5-7 days/week, at a BORG of 11-13 or HR of 111. He had no walking prescr ibed until had was able to bicycle everyday. Once he was able to tolerate bicycling ever yday, the prescription was to add 15 minutes of walking a day as tolerated. Fina lly, subject six (6) wa s the only non-married subject included in the visual analyses. Additionally, his BMI places him in the morbidly obese category. He also was the only one with significant hospitalizations during the entire st udy. He had two HF hospitalizations between discharg e and twelve weeks, one HF hospitalization between twelve and twenty four weeks, and one HF hospitalization between twenty four and fifty two weeks. Over the entire study, some important variable s changed for these subjects that were not captured visually in the RVA. Tables 4-5 and 4-6 are summary tables of some of these variables. Their NYHA classification, V O 2max and total exercise time (TET) are some of the original studys measures of the subjects functional capacity. Medications are the number of HF related medications the subject was taking. Assessments are the number of positive HF assessment symptoms (i.e. JVD, peripheral edema, S3) the su bject possessed. IV inotropes indicates if the subject received IV inotr opes while in the hospital.
68 Trend Analysis Days exercised per we ek. The regression model determined that the amount of days exercised per week was the greatest predictor of variance within the m odel and subsequently a strong contributor to dail y weight variability. For subject one, the best tr end fit for the amount of da ys exercised was the cubic r2 ( r2=0.728) indicating an overall sin wave with an increasing trend. When viewing the first twelve weeks there is a separate trend from the second twelve weeks. The initial trend is downward and linear while the second twelve week s has a flat, linear trend that has only a mild upward trend. For subject two, best trend fit for the am ount of days exercised was the quadratic r2( r2=0.117) indicating a mild horseshoe trend with the peak occurring in th e mid to later portion of the time frame. Of note, there is only a mild mathematical difference between the two trends (linear r2=0.089). Subject two had four downward spik es of weeks with few days of exercise (weeks 4, 15, 26 and 39) that demonstrated a sub tre nd. In this sub trend, th ere is an increase in days exercised with time. There is no clear sepa ration between twelve and twenty four weeks. After twenty fours weeks there is more consistency and a more linear, flat trend. For subject three, the best trend fit for th e amount of days exercised was the quadratic r2 ( r2=0.475) indicating an overall arch ing effect with an increasing trend. There is only a mild mathematical difference betw een the two trends (linear r2=0.439) with th e linear trend demonstrating a linear, upward trajectory. The fi rst twelve weeks demonstr ate a large variability that begins to even out at week eight. At tw elve weeks the variability has become stable, but there was a downward trend betw een weeks 16 and 20. During the maintenance phase the trend is to be in the high range and to be stable in the variability through the remainder of the study. For subject four, the best trend fit for the amount of days exercised was the cubic r2 ( r2=0.269) indicating a mild sin wave effect with a slight increasing trend. Th e trend fits the data
69 accurately as the subject is steadily exercisi ng with weeks of higher than normal dispersed evenly throughout. Then the subj ect returns to a steady three days per week of exercise with no spikes below three days per w eek in the maintenance phase. For subject five, the best trend fit for th e amount of days exercised was the cubic r2 ( r2=0.163) indicating a mild sin wave effect with a slight decreasing tre nd. The trend fits the data accurately as the subject is steadily exercising with weeks of lower than normal dispersed throughout and then a return to a steady five days per week and no spikes above five days per week after week 2. For subject six, the best trend fit for th e amount of days exercised was the cubic r2 ( r2=0.251) indicating a horseshoe ef fect with the peak occurring between weeks 16 and 28. The subject performed no exercise in the first twelve weeks, performe d intermittent exercise in the second twelve weeks, and perfor med intermittent exercise as th e study concluded with a trend downward. Minutes exercised each session. The regression model determined that the amount of time exercised during each exercise session was the ne xt greatest predictor of variance within the model. For subject one, both the cubic and lin ear fit produced a similar fit (cubic r2=0.181, linear r2=0.171) with the trend being a fl at, relatively linear increasing trend fo r the amount of time exercised each session. The subjec t showed greater variab ility in the first tw elve weeks than in the second twelve weeks. Regard less of which twelve weeks, both demonstrated separate linear increasing trends. For subject two the best trend fit for the am ount of time spent exercising each session was the cubic r2 ( r2=0.331) indicating a horseshoe effect w ith the peak occurring between weeks 8
70 and 30. The subject demonstrated an increasing, linear trend in the fi rst twelve weeks that became flat and remained linear in the second tw elve weeks. During the maintenance phase, the trend became linear and decreasing as the subject began exercising less each session. For subject three, the best trend fit for th e amount of time spent exercising each session was the cubic r2 ( r2=0.331) indicating a sin wave with an increasi ng trend. The first twelve weeks demonstrated great variability until week 8, at which point the time increased steadily with less variability until weeks 20 to 24. During the maintenance phase the subject exercised steadily demonstrating a flat linear trend. For subject four, both the quadratic and lin ear fit produced a similar fit (quadratic r2=0.05, linear r2=0.003) with the trend being a flat, re latively linear, upside down horseshoe trend for the amount of time exercised each sessi on. The subject demonstrated a great deal of variability during the first twelve weeks and then stabilized during the se cond twelve weeks at 30 minutes each session. The maintenance phase had few data points that lead to a flat, even trend. For subject five, the best trend fit for th e amount of time spent exercising each session was the cubic r2 ( r2=0.176) indicating a soft sin wave with an incr easing trend. The subject had a strong, increasing linear tr end in the first twelve weeks and a mild, downward linear trend in the second twelve weeks. The maintenance phase produced a flat, even line ar trend at 45 minutes each session. For subject six, the best trend fit for the am ount of time spent exercising each session was the cubic r2 ( r2=0.197) indicating a horseshoe effect with a peak at weeks 28 to 34. The first twelve weeks have a flat, linear trend at 0 minutes exercised. The second twelve weeks demonstrate a downward trend due to the 12 week time of 60 minutes. The maintenance phase
71 demonstrates erratic activity during weeks 24 to 36 and then no activity after week 36 until week 52. Exercise intensity during each session. The model did not dete rmine that intensity during each session was a signi ficant predictor of varian ce within the model. For subject one, the best tre nd fit for the amount of time spent exercising each session was the cubic r2 ( r2=0.273) indicating soft sin wave that is flat. The first twelve weeks has a very slight increasing linear trend. The second twelve week s have a flat linear trend. For subject two, the best trend fit for the amount of time spent exercising each session was the cubic r2 ( r2=0.592) indicating a sin wave with an increasing trend. The first twelve weeks have a strong, increasing linea r trend. The second twelve week s have a flat, linear trend. The maintenance phase holds flat at an intensity of a BORG of 11. For subject three, both the qua dratic and linear fit produced a similar fit (quadratic r2=0.021, linear r2=0.011) with the trend bei ng a flat, relatively linear, upside down horseshoe trend for the intensity exercised each session. Th e first twelve weeks de monstrate a decreasing linear trend while the second tw elve weeks demonstrate an in creasing linear trend. The maintenance phase demonstrates a linear, flat trend with a few spikes of higher intensity. For subject four, the best tr end fit for the amount of time spent exercising each session was the cubic r2 ( r2=0.353) indicating soft sin wave with decreasing trend. The first twelve weeks trend is flat due to the er ratic nature of the subjects exer cise intensity. The second twelve weeks even out at a BORG of 11 and produces a flat, linear trend. Maintenance phase demonstrates a linear, downward trend. For subject five, both the cubic and lin ear fit produced a similar fit (cubic r2=0.092, linear r2=0.02) with the trend being a flat, relatively linear, upside down horseshoe trend for the
72 intensity exercised each session. The first twelve weeks produce a flat linear trend at a BORG of 12. The second twelve weeks pr oduce a strong, decreasing linear tr end. The maintenance phase has a great deal of variability and pro duces a mild increasing linear trend. For subject six, both the c ubic and linear fit produ ced a similar fit (cubic r2=0.288, linear r2=0.279) with the trend being a flat, relatively linear, in creasing trend for the intensity exercised each session. The subject did not exercise in th e first twelve weeks ther efore did not yield any data for the first twelve weeks. The second tw elve weeks demonstrate a mild, increasing linear trend. The maintenance phase also demons trates a mild, increasing linear trend. Variability Analysis Daily we ight variability. Within the multiple regression, daily weight variability was assessed using the standard deviation. To assess an individuals vari ability graphically, two graphs were created. A standard error bar graph was cr eated which is a plot of the actual daily weights for each week with the standard error graphically displayed as well as the median. Additionally, a trend of the sta ndard deviation over the durati on of the study was plotted to develop a visual sense of the variability over time. Standard error. For subject one, the st andard error of daily we ight was greatest during the first twelve weeks of the study. During the second twelve week s there was a marked decrease in the standard e rror. Additionally, his mean we ight in pounds dropped during the second twelve weeks. For subjec t two, the standard error of daily weight was greatest during the first four weeks of the study. There was also a small peak in variability during weeks 20 to 22. Additionally, his weight in pounds increased at week 20 and remain ed higher until the end of the study. For subject three, the standard error of daily weight was grea t during the first half of the study. Variability did not begin to decrease until week 28 at wh ich point variability decreased
73 and his weight decreased as well. For subject four, the standa rd error of daily weight was greatest during the first eight week s of the study with a spike in va riability again at weeks 24 to 41. The subject had a jump in mean weight at we ek eight and then a decrease in variability at week 10 that maintained until week 24. His wei ght in pounds did not return to discharge values. For subject five, the standard error of daily weight was greatest during baseline with small spikes at 15 to 24 weeks. The variability was consistent as was the increase in weight in pounds. This subject was underweight according to his baseline BMI of 15.5. For subject six, the standard error of daily weight was great during the entire study. At some points the variability was too great to include on the graph a nd keep representative with the other subjects. He had three weeks of small decreases in variability during weeks 1, 12, and 30. Standard deviation. For subject one, both the cubic and linear fit produced a similar fit (cubic r2=0.246, linear r2=0.206) with the trend being a relati vely linear, downward trend for the standard deviations of daily weig hts or daily weight va riability. The first twelve weeks starting at week one, demonstrate a strong, increasing linear trend. The second twelve weeks demonstrate a mild, downward linear trend wi th variability values around or below one. For subject two, the best trend fit for da ily weight variability was the quadratic r2 ( r2=0.812) indicating an upside down horseshoe trend. The first twelve weeks starting at week one, produced a sharp, downward, linear trend. The second twelve weeks produced a flat, linear trend that increased very mildly. The maintenan ce phase also produced a fl at, linear trend that maintains with values below one. For subject three, the best trend fit for daily weight variability was the cubic r2 ( r2=0.409) indicating soft sin wave with a fl at trend. The first twelve weeks demonstrated a mild increasing,
74 linear trend. The second twelve weeks demonstrated a mild decreasing, linear trend. The maintenance phase held a flat, linear tr end with values maintaining around one. For subject four, the best trend fit for daily weight variability was the cubic r2 ( r2=0.324) indicating an upside down horseshoe trend. The first twelve weeks produ ced a mild increasing, linear trend. The second twel ve weeks produced a mild decreasing, linear trend. The maintenance phase produced a mild increasing, lin ear trend with values ranging from 0 to 3. Both subjects five and six had extreme base line variability values that made trending their daily weight variability a challenge. C onsequently, for both subjects a graph with and without baseline values is included for analysis The graphs without baseline values will be textually summarized due to relevan ce of trend related to exercise. For subject five, the best trend fit fo r daily weight variability was the cubic r2 ( r2=0.216) indicating a soft sin wave effect with an increasing trend. The first twelve weeks demonstrated flat, linear trend that decreased mildly. The second twelve weeks demonstrated a dramatic increasing, linear trend. The ma intenance phase demonstrated a linear, decreasing trend with values ranging from 3.5 to 0.45. For subject six, the best trend fit for daily weight variability was the cubic r2 ( r2=0.406) indicating sin wave effect with an increasing trend. The firs t twelve weeks produced a flat, linear trend at 1.7. The second twelve weeks produced a sharp increasing, linear trend. The maintenance phase produced a mild, increasing linear trend with values that ranged from 2 to 8. Level of Change Analysis Level of change analysis is unique to SSM and the assessment of an interventions effectiveness from one phase to another. Alt hough for this study an intervention is not being assessed, the effectiveness of time on the variables of interest and how the original studys intervention may or may not have had an effect on these variables was of interest. Therefore,
75 baseline to start of intervention was the first level of change interval assessed. The second interval selected was the twelve week to twenty four week as when doing the visual analysis on an outlier it was noted that the bicycle interventi on may or may not have been a factor in the dissertation study. It was at tw elve weeks that the bicycle wa s introduced to those who could tolerate the change in protocol. Finally, twenty four weeks to the exit of the study was chosen to view if there was a maintenance phase or a washou t of the intervention. W ith exercise, there is an individual choice to mainta in or to cease the program. Days exercised per week. At each time point the amount days the subject exercised on that given week is the value plotted. The amount of change is then assessed for each interval. Subject one changed a total of 4 days from ba seline to week one of the study, a total of one day from start of the study (w eek 1) to the mid-point of th e study (week 12), a total of one day from mid-point (week 12) to the end of the intervention (week 24), and had a grand total of 6 days of change from baseline to the time he exited the study. Subject two changed a total of 5 days from baseline to week one of the study, ha d no change from start of the study (week 1) to the mid-point of the study (week 12), had no change from mid-point (week 12) to the end of the intervention (week 24), had no change from the e nd of the intervention to the exit of the study, and a had grand total of 5 days of change fr om baseline to the tim e he exited the study. Subject three had no change from baseline to week one of th e study, a total of three days of change from start of the study (week 1) to th e mid-point of the study (w eek 12), a total of two days from mid-point (week 12) to the end of th e intervention (week 24), had no change from the end of the intervention to the exit of the study, and had a gran d total of 5 days of change from baseline to the time he exited the study. Subject f our changed a total of 2 days from baseline to week one of the study, a to tal of one day from start of the stud y (week 1) to the mid-point of the
76 study (week 12), a total of three days from mid-point (week 12) to the end of the intervention (week 24), dropped two days from the end of the intervention to the exit of the study, and had a grand total of 6 days of change from baseline to the time he exited the study. Subject five changed a total of 7 days from baseline to week one of the study, dropped a total of two days from start of the study (week 1) to the midpoint of the study (week 12), had no change from mid-point (week 12) to the end of the intervention (week 24), had no change from the end of the intervention to th e exit of the study, and had a grand total of 7 days of change from baseline to the time he exited th e study. Subject six had no change from baseline to week one of the study, a total of one day of change from star t of the study (week 1) to the mid-point of the study (week 12), had no change from mid-point (week 12) to the e nd of the intervention (week 24), dropped one day from the end of the interven tion to the exit of the study, and had a grand total of 1 day of change from base line to the time he exited the study. Minutes exercised each session. At each time point the average amount of minutes the subject exercised each exercise session that given week is the value plotted. The amount of change is then assessed for each interval. Subject one changed a total of 22 minutes from baseline to week one of the study, a total of 8 minutes from start of the study (week 1) to the mid-point of the study (week 12), had no change from the mid-point (week 12) to the e nd of the intervention (week 24), and had a grand total of 30 minutes of change from baseline to the time he exited the st udy. Subject two changed a total of 16 minutes from baseline to week one of the study, changed 24 minutes from start of the study (week 1) to the mid-poi nt of the study (week 12), had no change from mid-point (week 12) to the end of the intervention (week 24), drop ped 20 minutes from the end of the intervention
77 to the exit of the study, and a had grand total of 40 minutes of cha nge from baseline to the time he exited the study. Subject three had no change fr om baseline to week one of the study, a total of 30 minutes of change from start of the study (week 1) to the mid-point of the study (week 12), a drop of 12 minutes from mid-point (week 12) to the end of the intervention (week 24), a total of 12 minutes of change from the end of the intervention to the exit of the study, and had a grand total of 30 minutes of change from baseline to the time he exited the study. Subject four changed a total of 30 minutes from baseline to week one of the stud y, had no change from start of the study (week 1) to the mid-point of the study (week 12), a drop of 10 minutes fr om mid-point (week 12) to the end of the intervention (week 24), had no change fr om the end of the intervention to the exit of the study, and had a grand to tal of 30 minutes of change from baseline to the time he exited the study. Subject five changed a total of 30 minutes from baseline to week one of the study, a total of 30 minutes from start of th e study (week 1) to the mid-point of the study (week 12), had a drop of 15 minutes from mid-point (week 12) to the end of the intervention (week 24), had no change from the end of the intervention to the exit of the study, and had a grand total of 60 minutes of change from baseline to the time he exited the study. Subject six had no change from baseline to week one of the st udy, a total of 60 minutes of change from start of the study (week 1) to the mid-point of the study (week 12), had a drop of 58 minutes from mid-point (week 12) to the end of the intervention (week 24), a drop of 3 minutes from the end of the intervention to the exit of the study, and had a grand total of 60 minu tes of change from baseline to the time he exited the study.
78 Exercise intensity during each session. At each time point the average intensity the subject exercised each exercise session for that given week is the value plotted. The amount of change is then assessed for each interval. Subject one changed a total of 0.85 BORG from start of th e study (week 1) to the midpoint of the study (week 12), had no change from the mid-point (week 12) to the end of the intervention (week 24), and had a grand total of 0.85 BORG of cha nge from baseline to the time he exited the study. Subject two changed a total of 6.2 BORG from start of the study (week 1) to the mid-point of the study (week 12), dropped 2.8 BORG from mid-point (w eek 12) to the end of the intervention (week 24), cha nged 0.6 BORG from the end of the intervention to the exit of the study, and a had grand total of 6.2 BORG of change from baselin e to the time he exited the study. Subject three had no values for the first 12 week s of the study. He had a total change of 0.8 BORG from the mid-point (week 12) to the end of the intervention (week 24), a drop of 1 BORG from the end of the interv ention to the exit of the study, and had a grand total of 1 BORG of change from baseline to the time he exited the study. Subject four changed a total of 2 BORG of change in a drop from start of the study (week 1) to the mi d-point of the study (week 12), a change of 0.8 BORG from mid-poi nt (week 12) to the end of th e intervention (week 24), a drop of 2.8 BORG from the end of the intervention to the exit of the study, and had a grand total of 4 BORG of change from baseline to the time he exited the study. Subject five changed a total of 1 BORG of change from star t of the study (week 1) to the mid-point of the study (week 12), had a drop of 7 BORG from mid-point (w eek 12) to the end of the intervention (week 24), a cha nge of 6 BORG from the end of the intervention to the exit of the study, and had a grand total of 7 BORG of change from base line to the time he exited the
79 study. Subject six had data for neither the first 12 weeks of the study nor the last 28 weeks of the study. There was a total change of 3 BORG fr om mid-point (week 12) to the end of the intervention (week 24) with a grand total of 3 BO RG of change from baseline to the time he exited the study. Daily weight variability. At each time point the daily weight variability for each subject as the given weeks standard deviation for the da ily weight values and is the value plotted. The amount of change is then assessed for each interval. Subject one changed with a drop of 2.89 standa rd deviations from baseline to week one of the study, a total of 0.56 standa rd deviations from start of the study (week 1) to the mid-point of the study (week 12), dropped 0.65 standard deviations from th e mid-point (week 12) to the end of the intervention (week 24) and had a grand total of 3.08 st andard deviations of change from baseline to the time he exit ed the study. Subject two changed a total of standard deviations from baseline to week one of the study, changed standard deviations from start of the study (week 1) to the mid-point of the study (week 12) standard deviations fr om mid-point (week 12) to the end of the intervention (week 24), droppe d standard deviations from the end of the intervention to the exit of the st udy, and a had grand total of standard deviations of change from baseline to the time he exited the study. Subject three had a drop of 0.06 standard devi ations from baseline to week one of the study, a total of 0.62 standard deviat ions from start of the study (w eek 1) to the mid-point of the study (week 12), a drop of 0.71 standard deviations from mid-point (week 12) to the end of the intervention (week 24), a drop of 0.26 standard deviat ions from the end of the intervention to the exit of the study, and had a grand total of 0.97 standa rd deviations of change from baseline to the time he exited the study. Subject four dropped a to tal of 2.97 standard devi ations from baseline
80 to week one of the study, changed a total of 0.37 standard deviations from start of the study (week 1) to the mid-point of the study (week 12), dropped 0.73 standard deviations from midpoint (week 12) to the end of the intervention (week 24), change d 1.77 standard deviations from the end of the intervention to the exit of the study, and had a grand total of 3.33 standard deviations of change from baseline to the time he exited the study. Subject five dropped a total of 19.63 standard deviations from ba seline to week one of the study, dropped 0.34 standard deviations from start of the study (week 1) to the mid-point of the study (week 12), changed 2.99 standard deviations from mid-point (week 12) to the end of the intervention (week 24), dropped 2.7 standard deviati ons from the end of the intervention to the exit of the study, and had a grand total of 19.97 sta ndard deviations of change from baseline to the time he exited the study. Subject six droppe d 20.38 standard deviations from baseline to week one of the study, had no change from start of the study (week 1) to the mid-point of the study (week 12), changed 1.12 standard deviations from mid-point (week 12) to the end of the intervention (week 24), changed 2.65 standard deviat ions from the end of the intervention to the exit of the study, and had a grand total of 20.38 sta ndard deviations of change from baseline to the time he exited the study. Summary In conclusi on, the hierarchical multiple regression model demonstrated that exercise intensity, frequency, and duration, while controlling for event of study exit and baseline daily weight fluctuation, significantly predicted daily weight fluctua tions within a NYHA class III/IV heart failure population after medical optimization (R2 linear= 0.713, F=3.224, p=0.015). The retrospective visual analysis yielded that the subjects pooled as responders, non-responders, and outliers behaved as the model predicted and yielded additional information. The type of exercise
81 performed may or may not be a f actor in daily weight variability. Other individual information was useful for hypothesis generati ng future research questions.
82 Table 4-1. Baseline characteris tics: Admission to discharge All Subjects N=54 Adm D/C Drop Between Weight (Pounds) 209.10 202.40 (6.70) Mean (SD) (48.40) (50.40) Systolic BP 114.70 101.60 (13.10) Mean (SD) (20.70) (14.60) Diastolic BP 68.10 63.00 (5.10) Mean (SD) (11.50) (12.40) PAR-7 Day recall 39.90 Mean (SD) (12.20) V O 2Max (L/Min) 12.1 Mean (SD) (3.4) Total Exercise Time (secs) 245.2 Mean (SD) (78.6) NYHA Class 0 6 (11%) 6 (11%) II 26 (48%) 42 (78%) 16 (30%) III 28 (52%) 6 (11%) -22(-41% IV Jugular Venous Distention 40 3 (37.00) 73% 5% -68% Peripheral Edema 28 6 (22.00) 51% 11% -40% Murmur 27 8 (19.00) 49% 15% -34% Cardiac S3 39 24 (15.00) 71% 44% -27% Rales 14 4 (10.00) 26% 7% -19% Enlarged Liver 9 2 (7.00) 17% 4% -13% Abdominal Edema 5 2 (3.00) 9% 4% -5% Cardiac S4 1 1 2% 2%
83Table 4-2. Correlation summary table Daily Weight Variability Days/ Weeks Mins/ Session Intensity Baseline Variability Exit Event Daily Weight Variability Bivariate -0.151, NS -0.031, NS -0.083, NS 0.448, p=.026 -0.107, NS Partial -0.404, p=.037 -0.289, NS -0.156, NS Days/ Weeks Bivariate -0.151, NS 0.798, p=.000 -0.22, NS 0.067, NS -0.103, NS Partial -0.404, p=.037 0.782, p=.000 -0.217, NS Mins/ Session Bivariate -0.031, NS 0.798, p=.000 -0.224, NS -0.1, NS -0.289, p=.030 Partial -0.289, NS 0.782, p=.000 -0.195, NS Intensity Bivariate -0.083, NS -0.22, NS -0.224, NS 0.099, NS 0.140, NS Partial -0.156, NS -0.217, NS -0.195, NS Baseline Variability Bivariate 0.448, p=.026 0.067, NS -0.1, NS 0.099, NS 0.398, p=.003 Exit Event Bivariate -0.107, NS -0.103, NS -0.289, p=.030 0.140, NS 0.398, p=.003
84 Table 4-3. Secondary variables correlation summary Pedometer reading PAR Time/ SessionDays/ Week Intensity Pedometer reading Bivariate 0.152, NS 0.434, p=.002 0.404, p=.004 -0.025, NS Partial 0.131, NS 0.565, p=.008 0.614, p=.003 -0.237, NS PAR Bivariate 0.152, NS 0.257, NS 0.156, NS -0.051, NS Partial 0.131, NS 0.323, NS .0153, NS -0.088, NS
85 Table 4-4. Summary of individual characteristics Subject 1 Subject 2 Subject 3 Subject 4 Subject 5 Subject 6 Characteristic Gender Male Male Male Male Male Male Age (years) 53.3 63.6 52.1 49.8 71.5 30.3 Race White White White White White White Marital Status Married Married Married Married Married Single Original Grp Assignment Control Control Exerci se Control Exercise Control Smoking Status Past/ Current Past/ Current Past/ Current Past/ Current Past/ Current NonSmoker BMI / Classification 31.9 / Obese 38.7 / Obese 30.7 / Obese 28.6 / Overweight 15.5 / Underweight 42.1 / Obese Etiology Ischemic Idiopathic Ischemic Idiopathic Idiopathic Idiopathic Mths Diagnosed 229 57 12 44 192 84 Ejection Fraction 10% 30% 20% 15% 15% 10% Cardiac Output N/A 4.7 4.5 3.34 N/A 5.2 HF Hosp in prior 6 mths 4 0 1 0 0 4 Co-morbidities HTN Yes Yes No Yes No No DM Yes Yes No No No No COPD Yes No No No No No Hypercholersteremia Yes Yes Yes No No No Kidney Disease No Yes Yes No No No Atrial Fib No No No Yes No No Thyroid Disease No No Yes No No No Past Conditions Cancer No No Yes No No No CABG Yes (2) No Yes (1) No No No MI Yes No Yes No No No AICD No No Yes Yes No No Pacemaker No No Yes Yes No No V-Tach/Fib No No Yes Yes No No Valvular Disease No No Yes No No No
86Table 4-5. Summary of subjects 1, 2 and 3 time variables Subject 1 Subject 2 Subject 3 Variable Adm D/C 12 Wk 24 Wk Adm D/C 12 Wk 24 Wk 52 Wk Adm D/C 12 Wk 24 Wk 52 Wk NYHA III III III IV III II II II II III III III II III Assessments 4 1 4 4 3 1 2 1 0 1 0 2 2 1 Medications 7 No 8 In Pt 4 8 7 No No 6 6 No 7 No IV Inotropes Yes N/A N/A Yes No N/A N/A N/A N/A Ye s N/A N/A N/A N/A VO2 Max (L/min) 12 11.8 10.5 11.2 14.8 15.7 8.4 8.7 8.7 9.9 TET (secs) 237 343 213 220 323 218 225 253 282 267 Table 4-6. Summary of subjects 4, 5 and 6 time variables Subject 4 Subject 5 Subject 6 Variable Adm D/C 12 Wk 24 Wk 52 Wk Adm D/C 12 Wk 24 Wk 52 Wk Adm D/C 12 Wk 24 Wk 52 Wk NYHA III II II II II IV IV III II II IV III III IV IV Assessments 3 1 0 0 0 4 0 0 0 0 4 1 0 2 0 Medications 8 No No No No 6 8 No No No 4 6 No No No IV Inotropes Yes N/A N/ A N/A N/A Yes N/A N/A N/A N/A Yes N/A N/A Yes N/A VO2 Max (L/min) 17 22 21.8 19.2 14 19.8 22.8 20.2 13.5 15.5 16.5 14.3 TET (secs) 421 552 564 485 182 294 309 326 274 317 346 173
87 Figure 4-1.Trend analysis for days exercised per week: Responders Figure 4-2.Trend analysis for days exercised per week: Non-responders
88 Figure 4-3.Trend analysis for days exercised per week: Outliers Figure 4-4.Trend analysis for amount of time sp ent exercising each session: Responders
89 Figure 4-5. Trend analysis for amount of time sp ent exercising each session: Non-responders Figure 4-6. Trend analysis for amount of tim e spent exercising each session: Outliers
90 Figure 4-7.Trend analysis for exercise in tensity during each session: Responders Figure 4-8. Trend analysis for exercise inte nsity during each sessi on: Non-responders
91 Figure 4-9. Trend analysis for exercise intensity during each session: Outliers Figure 4-10.Variability analysis of daily weight: Responders
92 Figure 4-11. Variability analysis of daily weight: Non-responders Figure 4-12. Variability analysis of daily weight: Outliers
93 Figure 4-13.Variability trend analysis for daily weight variability: Responders Figure 4-14. Variability trend analysis fo r daily weight variability: Non-responders
94 Figure 4-15. Variability trend analysis for daily weight variability: Outliers
95 Figure 4-16. Level of change analysis fo r days exercised per week: Responders Figure 4-17. Level of change analysis for days exercised per week: Non-responders
96 Figure 4-18. Level of change analysis fo r days exercised per week: Outliers Figure 4-19. Level of change analysis for amount of time exercise each session: Responders
97 Figure 4-20. Level of change analysis for amount of time exercise each session: Non-responders Figure 4-21. Level of change analysis for amount of time exercise each session: Outliers
98 Figure 4-22. Level of change analysis for exer cise intensity while exercising: Responders Figure 4-23. Level of change analysis for exerci se intensity while exercising: Non-responders
99 Figure 4-24. Level of change analysis for ex ercise intensity while exercising: Outliers Figure 4-25. Level of change analysis for daily weight variability: Responders
100 Figure 4-26. Level of change analysis for daily weight variabi lity: Non-responders Figure 4-27. Level of change analysis for daily weight variability: Outliers
101 CHAPTER 5 DISCUSSION Discussion of Findings The dissertation study, coupled with the baseline demographic results from the original study, provided hypothesis generating, thought pr ovoking findings. The study sam ple was a convenience sample which when compared to a similar population (ADHERE registry, Yancy, et. al., 2006) was similar in gender distribution (study = 77% male, ADHERE =60% male), race distribution (study = 83% white, ADHERE = 78% wh ite), and age (study mean age= 57.8 years, ADHERE mean age = 69.8 years). Although the study sample was not completely representative of the disease population, the sample was small (n=54) and these variations should not theoretically affect the physiology outcomes or conclusions of the study. Summary of Data Analyses The main bivariate relationships of interest were daily weight variability and the exercise prescription variables (exercise frequency, duration, and intensity). A moderate, statistically significant relationship that emerged was th at as the number of days exercised (r=-0.404, p=0.037) and the amount of time exercised each session ( r =-0.289, p=NS) increased, the daily weight variability decreased. This indirect relationship demonstrat ed that as subjects increased their overall exercise training the physiological response was to decr ease the fluctuations in the daily weight seen in a 7 day time frame. Th e bivariate relationship supported the proposed physiological model. A second strong, statistically significant rela tionship existed between the amount of time exercised and the amount of days exercised ( r =0.782, p<.000). Initially, multicollinearity was of concern, yet due to the separateness of the variables individually, it di d not prove to be an issue.
102 Ultimately, it can be interpreted that, as a group, the individuals who exercised more frequently also were able to tolerate ex ercise at longer intervals. Summary of multivariate analyses findings. Overall, the hierarchical multiple regression model was predictive as theoretically indicated. Th e model predicted 20% of the variance, yet, with 80% of the variance unaccounted for, there is an indica tion that exercise alone is not enough of a down regulati on of the sympathetic nervous system to counter daily weight fluctuations. Additionally, the lack of effectiveness of exercise intensity within the model to be individually predictive of variance may stem fr om the indirect relationship between the amounts of days exercised and the intensity at which the subjects exercised ( r =-0.217, p=NS). This indirect relationship indicates that within the advanced heart failure population as subjects exercised more frequently and/or for greater length s of time, they either tolerate it at a lower intensity or developed a traini ng effect. Although the overall model with the full exercise prescription was predictive and i ndicates that exercise can and s hould be used as an adjunctive medical management application to this patien t population, further res earch is indicated to understand the exercise prescription within this patient population. Summary of retrospective visual analyses findings. Retrospective visual analysis was used to conduct a more in depth assessment of how responders, non-responders, and outliers profiled in response to exercise a nd fluid instability. This type of visual analysis is new and a hybrid version of the traditional visual analysis used in SSM. Due to the innovative application of this analysis, a set inclusion/exclusion criteria was used, as not all subjects included in the original study were expected to have had data detailed enough to attain the information necessary for sufficient visual analysis. Additionally, a detailed subject profile, as a descriptive data summary, was developed for each analysis as is traditional within SSM. The retrospective visual
103 analyses fell in line with the statistical conclusions of the mu ltiple regression with responders being responders, non-responders being non-re sponders, and outliers being outliers. The most significant piece of information gathered from RVA was that the type of exercise performed may or may not have contri buted to the daily weight variability profile demonstrated. Past research has demonstrated th at the vasculature of HF patients is damaged (Nakamura, 1999; Nakamura, et al, 1995; Nakamura, et al, 1996; Nakamura, et al, 1997; Wray, et. al., 2007). There is the possi bility that the damaged vasculat ure requires a more direct mode of lower body strengthening (Braith & Beck, 2008) to create shear st ress to create the mechanism of action seen for maintenance of fluid balan ce within this study. A much better understanding of the possible differences between the two type s of aerobic training (wal king vs. bicycling) is warranted as it is current practi ce to use the two aerobic protocols interchangeably within this population. Additional information beyond the fact that th e RVA supported the sta tistical findings of the multiple regression was determined when doing the individual analyses. When analyzing the data visually, it was found that the subjects who used a bicycle went from responder profiles to non-responder profiles when the bicycle protoc ol was introduced. Additionally, a subject who participated in cardiac rehabil itation and was selected for RVA, had a paradoxical response while in rehabilitation yet once rehabilitation was completed, demonstrated a responder profile. Additional features drawn from the RVA re volve around the quality and amount of data generated for analysis. Although there was not su fficient data to take the multiple regression model to the year mark, RVA subjects offered ri ch data until the year mark. Also, individual subject characteristics gathered from the RVA generated hypothese s and refined ideas regarding
104 protocols for future research. Ther e were no other distinct group tre nds gathered in regards to the studys research question. Relevance and implications of findings Accounting for 20% of the variance, the exercise prescription, as determined by exerci se frequency, duration, and intensity, is an excellent adjunctive therapy for advanced heart failure patien ts. Beyond improving quality of life, VO 2 max, or functional capacity, exercise can also decr ease daily weight fluc tuations that often plague the advanced heart failure patient. These fluctuations can and often do lead these patients in a downward spiral of a hemodynamic cr isis and subsequent hospitalization. Retrospective visual analysis was able to yield data that not only supported the regression model yet was hypothesis generating regarding the 80% of unaccounted for variance. The unaccounted variance may not only be variance between subjects on pharmacology, nutrition, and/or medical status but may also have been on th e type of exercise the s ubjects participated in. These findings were only availabl e with visual analysis. Sympathetic Nervous System, Fluid Bala nce Instability, and Exercise Training The physiological fram ework for this study was founded on the comprehensive, consensus model designed by the author. The research th at was the foundation for this model was begun as early as the 1970s. The overa ll thought process that prope lled this study was that the overactivation of the sympathetic nervous system of the HF patient places the HF in a fluctuation between homeostasis and non-homeostasis. The consta nt fluctuation places the patient in a state of fluid balance instability via the SNS increas ing the levels of AVP and activating the RAAS which subsequently increases the leaves of angi otensin II (Bekheirnia & Schrier, 2006; Berl, et. al., 1979; Cadnapaphornchai, et. al., 2001; Gademan, et. al., 2007; Hensen, Abraham, Durr, &
105 Schrier, 1991; Schrier, et. al ., 1999; Selektor & Weber, 2008; Wa ng et al., 2001). These actions lead to the symptomatic expression of edema and daily weight fluctuations. Exercise training has been shown to decrease sympathetic nervous system activation via decreased NE levels, decreased RHR, increase d HR variability and PHR response, decreased AVP, angiotensin II, and aldoster one (Braith, et. al., 1999; Cider et al., 1997; Gademan, et. al., 2007; Keteyian et al., 1999; Kiil avuori, et. al., 1999; Larsen et al., 2004; Nishiyama et al., 2006; Passino et al., 2006; Selig et al., 200 4; Tyni-Lenne et al., 1999). Ex ercise training has yet to be shown to modify symptomatic expression of fluid balance instability. The dissertation statistical findings demonstrat ed that exercise decreased daily weight fluctuations which are a sympto matic expression of the over ac tive sympathetic nervous system of the HF patient. The RVA findings further supported the statistical findings and added depth to the individual variability of the findings. Overall, there was not a clear physiological link demonstrated due to no physiological variables such as NE, AVP, aldoste rone, or angiotensin measures being taken. Yet, with very little invasi on to the subject, exercise demonstrated a clear symptomatic alleviation of the over activation of the SNS. Endothelial Changes, Fluid Balance Instability, and Exercise Training The HF patient has ma ny neural hormonal eff ects related to the syndrome. Researchers and clinicians alike are unsure as to the exact order of progression but are su re that there is a downward spiral once onset occurs. One aspect of the syndrome that has demonstrated great potential as a start point is the endothelial dysfunction that a majority if not all patients demonstrate. The dysfunction has as its most ba sic element a decrease in bioavailability of NO (Bauersachs & Schafer, 2004). The lack of NO l eads to a multitude of consequences all of which leave the patient with a dysfunctional vasculature unable to distribute oxygen to the organs and muscles of the body (Adachi, et al., 1997; Agostoni & Bussotti, 2003; Maguire,
106 Nugent, McGurk, Johnston, & Nicholls, 1998; Yoshida, Nakamura, Akatsu, Arakawa, & Hiramori, 1998). Exercise increases the bioavailability of NO to the endothelium of HF patients via increasing the sheer stress placed on the endothe lium. The sheer stress then increases eNOS production and antioxidant up regulation which decreases ROS subsequently also increasing NO (Bauersachs & Schafer, 2004; Enne zat, et. al., 2001; Froelicher & Myers, 2006). The upstream effect of exercise on the vascul ature has a downstream of effect on the over activation of the SNS by decreasing the over activation signaling fr om the low NO, decreasing the activation of the RAAS, and may even have an effect on the intra cellular to ex tra cellular movement of fluid that has yet to be explored. The retrospective visual analys is yielded one group finding that was of great significance. Three of the six, or 50%, subjects demonstrated a mode of exerci se effect. The type of aerobic exercise the subjects performed determined wh ether the subject demonstrated a responder or non-responder profile. The treadmill, which has a load or weight bearing and muscle pump mechanism to the aerobic component, had a stro ng responder profile whereas the bicycle, which does not have a load or weight bearing component and minimal to no muscle pump mechanism due to extended plantar flexion, did not have a responder profile. Th e results from the RVA findings of the dissertation study lend credence to the proposal that the e ffects of exercise for fluid balance instability are fu rther upstream within the model and require a sheer stress threshold. The threshold was met when subjects were walking and it was not met when subjects were riding the bicycles. Summary In conclusion, the statistical analyses de monstrated the underlying potential of the exercise prescription and subsequent hom e based exercise training program to affect the fluid
107 status balance of NYHA class III /IV HF patients. The retrosp ective visual analysis added a greater depth of understanding regarding the vari ability seen within th e individual subject. Together the combined data analyses were able to shed light on the potential of the exercise prescription to have a physiological effect on fluid status that would translate into a symptomatic expression of daily weight fluctuations. Not only were these analyses able to demonstrate such an effect, these analyses genera ted more hypotheses regarding the physiological role of exercise in the medical management of fl uid instability of NYHA Class III a nd IV heart failure patients. Although this study has limitations, it has demons trated that exercise is a successful adjunctive therapy to managing the daily weight variability or fluid stat us instability of NYHA Class III/IV patients. The fluid status instability is often a deb ilitating aspect of the syndrome of heart failure due to the hemodynamic fluctuat ions leading to acute decompensation and subsequent hospitalization. This study yields the first piece of evidence that exercise training or the exercise prescription can alter the effects of intra cellular to ex tra cellular fluid shifts seen in the heart failure (HF) patient. If one component of the exercise prescr iption has greater effect than another, was not determined. Neither was this study able to dete rmine a dose response. Overall, the study has developed a new set of hypot heses and generated more questions than it has answered. Yet, it is a true first step to demonstrating that exercise has an effect on fluid instability within the NYHA class III/IV patient p opulation and that further research needs to be conducted to truly under stand this effect. Limitations of the Study Methodological Limitations As a retrosp ective data analysis, there are lim itations to the questions that can effectively be asked of the data. The original study was a be tween group design, yet as an intention to treat study, there were confounding variable s for the research questions re lating to this study. These
108 issues made it impossible to keep the original between group design. The change in study design between the original design and the secondary analysis is th e greatest methodological limitation. To remedy this limitation, a randomized, prospective study would have been conducted with the same population and using the exerci se protocol on all participants. Statistical Limitations The greatest statistical limitation is the sam ple size. Although a pilot study and over 50 subjects for the original study was an acceptable amount, for a multiple regression 54 subjects is not enough for a robust model. For this particular model, which wa s statistically significant, the size ultimately did not hamper the end result. Yet, sample size may have been an issue in the determining of the beta-coefficients and which variables were significan t predictors of the variance. A larger sample would have been pref erable for the analysis that was performed. To remedy this limitation a power analysis would ha ve been conducted prio r to recruitment of subjects for a multiple regression. Then recruitment would have been aimed at 20% over the number need to attain sufficient power to account for possible attrition. Other Limitations The original study used self report for the daily activity l ogs and the dissertation study utilized these self -report logs substantially for th e analysis of the actual exercise prescription and the reporting of daily weight. As such, the reliability of the daily activity logs is skeptical. To address this limitation, the original PI used pedo meter readings (steps per days) to corroborate the findings of the daily activity l ogs. The strength in the pedomete r is that it is an objective measurement of physical activity whereas the weakness is that it does not measure actual exercise directly. The bivariate analysis dete rmined that as subjects both exercised more each session ( r =0.565, p=.008) and increased the amount of days they exercised each week ( r =0.614, p=.003) they took statistically signif icant more steps each week. These trends were parallel to
109 those demonstrated within the regression model i ndicating consistency between the exercise logs and the pedometers. As a retrospective analyses, there are a number of other limitations that the author was unable to control for nor able to account for within the statistical approach. Although the pharmacological variations for each subject throughout the study are known there was no systematic way to account for the subjects individual pharmacological regimen throughout the study and on each day. There was no system atic way to determine how influential pharmacological regimen was on an i ndividuals daily weight variability or to determine if the subject was medically optimized th roughout the entire study. As th e subjects had been medically optimized at the beginning of the study and we re successfully randomized for the original study, it was assumed that pharmacological regimen af fected subjects equally throughout the entire study. The original study did not collect any data rega rding the nutritional status or the nutritional habits of the subjects. The nutritional variations may have also affected the daily weight variability of the subjects, specifically sodium intake. Sodium intake, although significant in water retention, is important in that sodium is reabsorbed more readily when angiotensin II is high and that occurs when the RAAS is being activated. In heart failure, sodium intake becomes problematic when the RAAS is activ ated outside of homeostasis. Th erefore, nutritional intake of high sodium foods would be of issue when the body is out of homeostasis or when not being medically optimized. Also of concern is high fl uid intake. Again, this was not measured and would directly affect daily weight variability wh en the individual was not able to adequately remove the fluid through normal homeostatic m eans. Consequently, these nutritional intake
110 variables would theoretically be of greater significance in the si cker, less stable individuals and of less significance in the more stable subjects. Finally, there were a number of individual su bject variations that were not accounted for within the study. These individual variations included, but are not limited to, time of year the subjects participated in the study (i.e. summer vs winter), whether the subjects exercised indoors or outdoors (induction of sweating), and/or hormonal levels. Thes e subjects variations could have directly affected the daily weight variations and were not accounted for within the statistical model nor were these variables documented with in the original studys medical records. Overall, there were many limitations to this small, pilot dissertation study that limit the application of the findings. However, the limita tions were addressed wi th enough resolution and the findings were strong enough, to warrant the use of these findings in the proposal of future studies. Analyzing Data Using Statis tical and Visual Analyses The use of both statistical and retrospective visual data analysis enabled a depth of understanding of the data that neither me thodology al one was able to yield. The results set forth have the potential to lead to new knowledge and approaches for improving healthcare and patient outcomes through exercise interventi ons in a population that have often been thought too sick to exercise and too sick to benefit from exercise In traditional studies that employ the single subject methodology there is no need to exclud e patients who are curre ntly taking certain medications or who can not perform standardized rigid exercise protocols with intense or frequent assessment. By employing a new variation of SSMs visual analysis in an innovative format, the researcher was able to accommodate fo r these same issues that arose during the study due to intense, accurate documentation. Thus this multi-format statistical design has the
111 potential to capture unique information that a trad itional clinical trail with a large sample could not. The philosophy underlying traditional single subject methodology, as well as the language and methods of this type of research, di ffer substantially from tr aditional basic science research. By retaining traditi onal statistical analys es and adding a new hybrid visual analysis, the researcher was able to gather a more in depth look at the indivi dual results as well as retaining generalizability. Employing both types of data analyses is unique, innovative and was clearly warranted and beneficial in this situation. Overall, the results from this study allowed for a more thorough understanding of the patterns and trends in the physiological outcomes following a progressive, individualized exercise protocol within the heart failure patient population and how best to proceed with future research in this area. Implications for Nursing Science The implications for nursing science revolve arou nd four key aspects: th e ability to add to current theory and new theory development in ex ercise adherence, the ability to add to and develop a strong line of research regarding physiologi cal applications of exercise in heart failure, to aid in the development of an exercise protocol or prescription that th e nurse, advanced nurse practitioner, or other primary car e provider can provide directly to HF patients, and to contribute to the overall evidenced based care of the heart failure patient. Nursing theory The current findings have the potent ial to add to nursing theory via the exercise adherence avenue. Currently exercise adherence uses two main theories as the foundation for the exercise program interventi ons: Banduras Social Cognitive Theory (SCT) (Bandura, Adams, & Beyer, 1977) and SCTs of f-shoots: Banduras Theory of Self -Efficacy (Bandura, 1977) and Weiners Attributional Theory of Achievement Motivation (ATAM) (Weiner, 1985). These theories have many overla pping aspects. Current interventional nurse
112 researchers have combined the theories to de velop consensus theory driven interventions. Resnick is one such researcher who has combined the Theory of Self-Efficacy and ATAM to drive her Exercise Plus Program (Resnick, Magaziner, Orwig, & Zimmerman, 2002). The goal of this exercise intervention is to increase exercise and physical ac tivity in the elderly after a hip fracture and subsequent surgery. The dissertation study presented contributes to this body of research knowledge via the intervention used in the original study and the success of the homebased intervention within th e advanced HF population. The original study used many aspects found w ithin SCT and ATAM to be effective for overall home based exercise programs and for elderly exercise programs. The home-based exercise program used self-monitoring via daily logs and weekly telephone calls from the study coordinator. Both techniques have been shown to be effective operationa lizations of the selfregulation concept of SCT (Umstattd & Halla m, 2007). Additionally, the sample itself was primarily male (77%), a majority of the participants were not working, the exercise was both low intensity and not uncomfortable, and there were immediate physical benefits for this population. All of these characteristics ha ve been shown to have a high probability of success for home based exercise programs regarding participant adherence based on SCT (Dishman, Sallis, & Orenstein, 1985). Finally, as a telephone m onitored, home-based exercise intervention, researchers have found based on the principles of SCT, these modes of delivery to be the most effective in regards to adherence in the elderly population (King, 2001). Physiological applications The findings of the current study provide information yielding the foundations that exer cise has physiological benefits beyond those previously known. Previously, research has danced around the down regulation of the sympathetic over activation within the heart failure population via targeting ne ural hormonal regulatory markers of this down
113 regulation (i.e. norepinephrine, aldosterone, angi otensin II) (Braith, et al, 1999; Keteyian, et. al., 1999; Kiilavouri, 1999; Passino, et. al 2006; Tyni-Lenne, et. al., 1999; van Tol, et. al., 2006; ) and the physiological expression (i .e. heart rate, heart rate vari ability) (Cider, et. al. 1997; Keteyian et al., 1999; Larsen et al., 2004; Nishiyama et al., 2006). Th is previous research was the foundation of the physiological framework of this study. Current researchers have progressed further back down the physiologic path towards the endothelial dysfunction (Hornig, Maier, & Drexler, 1996; Kobayash i et al., 2003; Roveda et al., 2003) and nitric oxide bioavailability (Hambrech t et al., 1998) and how exercise effects these aspects. No one had yet to ve ntured down the path to the dire ct symptomatic expression of the sympathetic over activation. This study has pr ovided the first step in the foundation of a research program down that path. The findings demonstrated a direct impact on physiological factors as well as symptomatic expression demonstr ating both scientific and clinical relevance. The findings provide the first brick of a research trajectory that will open doorways to understanding how the body naturally manages or in some cases mismanages self-regulation that leads to severe illness in some individuals. Exercise prescription/protocol The presented study and its results get researchers one step closer in the development of a provider ba sed prescriptive home-based exercise program for all heart failure patients. Additionally, the findi ngs warrant a larger pr ospective study examining the effects of intensity and type of aerobic exercise more closely to determine the effect of these variables on daily weight variability more clea rly. There are still many questions such does exercise need to be a lifestyle change or can a specific period of exerci se be effective enough to decrease a HF patient from one NYHA class to another and sustain su ch decrease (i.e. dose dependent response)? Additionally, does severity of illness or personal characteristics contribute
114 more to the positive physiologic response of dail y weight fluctuations to exercise? Although these findings are additive to other studies from past research and ones currently being done, there is still much work to be done before a standard prescrip tion can be set out for all heart failure patients. Evidenced based practice. As a pilot study, this study ha s yielded beneficial knowledge for the HF patient, the cardiologist, the primary ca re provider, and the researcher alike regarding the care of the heart failure patient. This st udy demonstrated an overall clinical improvement in the patients condition via an improvement in one of the factors that contri butes to the recurrent discomfort in this patient populationfluid volum e overload via fluid shifts. Such a finding adds to the evidenced based practice resources needed for the best practice of our practitioners. Contributing to the improvement of nursing sc ience by contributing to the improvement of patients lives is the essence of a nurse researcher. Implications for Clinical Practice The findings have one primary clinical im pli cation: exercise trai ning in the home-based environment for the NYHA Class III/IV heart fail ure patients is both safe and effective in decreasing the fluid instability seen in these cong ested patients. A secondary take home message is that these patients exercised minimally (on average 13 minutes each session and only 3 days a week) at a low exertion level (between Light and Somewhat Hard). As a group, these were very sick patients, as eight died and eight had transplants before the year was over, yet only two had to be removed due to non-compliance to protocol. Consequently, for the advanced HF population a low intensity, low duration, at modera te frequency exercise program was sufficient to decrease daily weight variability. Additionally, when using a home based exerci se program, the use of SCT to guide the implementation of the program will increase the effectiveness of the adherence of the
115 participants. Promoting education and providi ng support via a comprehensive HF team provides self-efficacy for exercise and self-regulation for the HF patient. Such actions will increase the probability of the HF patient adhering to the pr ogram and consequently receiving the benefits of exercise. Recommendations for Future Research Physiological Model for Hemodynamic Fluc tuation Modulation in He art Failure by Exercise Training The author proposed a comprehensive, cons ensus physiological heart failure symptom model (Figure 2-1, page 38) which has many asp ects of research that have previously not communicated with each other. The unification of different branches HF research has benefits and challenges. The model pr ovides many fresh new approa ches to viewing exercise, endothelial dysfunction, and the ove r action of the sympathetic nervous system, or as others refer to neural hormonal activation (Gademan, et. al ., 2007), and their role in HF and the medical management of HF. The disadvantages revolve around teasing out nuances between disciplines and very specific areas of research. The re search does not always translate well between disciplines or between bench re search and clinical researc h. Regardless, the model has demonstrated potential and is worthy of further research and investigation. Exercise Training and Endothelial Dysfunction in Heart Failure related to Fluid Balance The dissertation study demonstrated that endo thelial dysfunction has the potential to be the strongest factor or player in regards to the physiological role of exercise within HF. The body of research supporting exercise as an agent of change for endothelial dysfunction is large, yet the mechanism of change is not definitive. Currently, the theory supports the action relating to the up regulation of NO in some manner. Th ere is the up regulation via sheer stress via eNOS and then there is the up regulation via ROS metabolites (i.e. anti-oxidation of ROS
116 therefore increasing NO bioava ilability) (Bauersachs & Schafer, 2004; Ennezat, 2001). Regardless of the mechanism the endothelial and neurohormonal players are many. Between endothelial-1, NO, ROS, EPC, e-NOS, RAAS, ANP, BNP, NE, and others yet to be discovered, the affect exercise has on these bio-markers and the subsequent hemodynamic measures are boundless. Further research needs to explore the interdependent roles to determine the true physiological role of exercise on heart failure and the maintenance of hemodynamic stability and fluid instability. Conclusions Heart failu re (HF) is a comp lex, often debilitating, chr onic disease that requires extensive, multidimensional as well as multi-disciplinary management. Research has demonstrated that exercise plays a key role in the prevention, as well as, the management of this disease through its abil ity to directly affect the many consequences or outcomes of HF. Within this study, a regression model demons trated that the full exercise prescription was predictive of fluid status instability and indi cates that exercise can and shou ld be used as an adjunctive medical management application to this patient population. The retrospe ctive visual analyses added depth to the statistical findings by both creating clarity, as well as, generating new hypotheses regarding the intra-individual and inter-i ndividual variability demonstrated. Together the data analyses yielded a much greater, holistic picture than either anal yses would have alone. These preliminary findings suggest that exercise can be a suc cessful adjunctiv e therapy to managing the daily weight variability or fluid st atus instability of NY HA Class III/IV patients that is often a debilitatin g aspect of the syndrome of heart fail ure. This enables nurse researchers to further develop lines of research regardi ng exercise adherence theory and to begin the foundational physiological theory for exercise as a modulator of the sympathetic nervous system within fluid volume overload. Additionally, and more importantly, these findings support the
117 growing body of evidence that the hemodynami c cascade that occurs prior to an acute exacerbation can be detected and even intervened upon. Exercise can play a role in the medical management, just as polypharmacy plays a key role, of this hemodynamic cascade. Additionally, this study further s upports the growing body of evidence that exercise may even be that key to unlocking the natural mechanisms of heart failure that have baffled practitioners and researchers alike for years. Continued research into the physiologic mechanisms while de termining at what intensity, frequency, duration, type, and length exercise is needed to be eff ective will not only prove to be beneficial for science but yield healthier subjects while doing so. Nu rsing researchers and practitioners can use this information to develop evidence ba sed practice interv entions regarding patient education, health promotion and exercise adherence. Ultimately, this dissertation study has yielded more questions than it has answered yet in doing so it has pointed researchers down wide open paths with unanswer ed questions to explore.
118 REFERENCES Abraham, W. (2007). Managing A DHF: Don't forget the hemodynamic goals. Paper presented at the Heart Failure Society of Amer ica Annual Meeting, Washington, D.C. Adachi, H., Nguyen, P. H., Belardinelli, R., Hunter, D., Jung, T., & Wasserman, K. (1997). Nitric oxide production during exerci se in chronic heart failure. Am Heart J, 134 (2 Pt 1), 196-202. Adamopoulos, S., Parissis, J., Kroupis, C., Georgiadis, M., Karatzas, D., Karavolias, G., et al. (2001). Physical training redu ces peripheral markers of inflammation in patients with chronic heart failure. Eur Heart J, 22 (9), 791-797. Adamopoulos, S., Parissis, J. T., Georgiadis, M., Ka ratzas, D., Paraskevaidis, J., Kroupis, C., et al. (2002). Growth hormone administrati on reduces circulating proinflammatory cytokines and soluble Fas/soluble Fas ligand sy stem in patients with chronic heart failure secondary to idiopathic dilated cardiomyopathy. Am Heart J, 144 (2), 359-364. Adamopoulos, S., Piepoli, M., McCance, A., Bernardi, L., Rocadaelli, A., Ormerod, O., et al. (1992). Comparison of different methods for assessing sympathovagal balance in chronic congestive heart failure secondary to coronary artery disease. Am J Cardiol, 70 (20), 1576-1582. Adams, V., Jiang, H., Yu, J., Mobius-Winkler, S., Fiehn, E., Linke, A., et al. (1999). Apoptosis in skeletal myocytes of patients with chroni c heart failure is associated with exercise intolerance. J Am Coll Cardiol, 33 (4), 959-965. Agostoni, P., & Bussotti, M. (2003). Exhaled ni tric oxide and exercise performance in heart failure. Arch Physiol Biochem, 111 (4), 293-296. Agostoni, P., Guazzi, M., Doria, E., & Marenz i, G. (2002). Pulmonary hemodynamic and tidal volume changes during exercise in heart failure. Ital Heart J, 3 (2), 104-108. Anand, I. S., Ferrari, R., Kalra, G. S., Wahi, P. L., Poole-Wilson, P. A., & Harris, P. C. (1989). Edema of cardiac origin. Studies of body water and sodium, renal function, hemodynamic indexes, and plasma hormones in untreated congestive cardiac failure. Circulation, 80 (2), 299-305. Anker, S. D., Chua, T. P., Ponikowski, P., Harri ngton, D., Swan, J. W., Kox, W. J., et al. (1997). Hormonal changes and catabolic/anabolic imba lance in chronic heart failure and their importance for cardiac cachexia. Circulation, 96 (2), 526-534. Anker, S. D., Clark, A. L., Kemp, M., Salsbury, C ., Teixeira, M. M., Hellewell, P. G., et al. (1997). Tumor necrosis factor and steroid me tabolism in chronic heart failure: possible relation to muscle wasting. J Am Coll Cardiol, 30 (4), 997-1001.
119 Anker, S. D., Volterrani, M., Pflaum, C. D., Stras burger, C. J., Osterziel, K. J., Doehner, W., et al. (2001). Acquired growth hormone resistance in patients with chronic heart failure: implications for therapy with growth hormone. J Am Coll Cardiol, 38 (2), 443-452. Armour, W., Clark, A. L., McCann, G. P., & Hillis, W. S. (1998). Effects of exercise position on the ventilatory responses to exercise in chronic heart failure. Int J Cardiol, 66 (1), 59-63. Bandura, A. (1977). Self-efficacy: toward a unifying theory of behavioral change. Psychol Rev, 84(2), 191-215. Bandura, A., Adams, N. E., & Beyer, J. ( 1977). Cognitive processes mediating behavioral change. J Pers Soc Psychol, 35 (3), 125-139. Bauersachs, J., & Schafer, A. (2004). Endothelial dysfunction in heart fa ilure: mechanisms and therapeutic approaches. Curr Vasc Pharmacol, 2(2), 115-124. Bekheirnia, M. R., & Schrier, R. W. (2006). Pathophysiology of water and sodium retention: edematous states with normal kidney function. Curr Opin Pharmacol, 6 (2), 202-207. Belardinelli, R. (1998). Muscle oxygenation ki netics measured by near-infrared spectroscopy during recovery from exercise in chronic heart failure. G Ital Cardiol, 28 (8), 866-872. Belardinelli, R., Georgiou, D., Cianci, G., & Pur caro, A. (1999). Randomized, controlled trial of long-term moderate exercise training in chronic heart fa ilure: effects on functional capacity, quality of life, and clinical outcome. Circulation, 99 (9), 1173-1182. Belardinelli, R., Zhang, Y. Y., Wasserman, K., Purcaro, A., & Agostoni, P. G. (1998). A fourminute submaximal constant work rate exercise test to assess cardiovascular functional class in chronic heart failure. Am J Cardiol, 81 (10), 1210-1214. Beniaminovitz, A., Lang, C. C., LaManca, J., & Mancini, D. M. (2002). Selective low-level leg muscle training alleviates dyspnea in patients with heart failure. J Am Coll Cardiol, 40(9), 1602-1608. Berl, T., Henrich, W. L., Erickson, A. L., & Schrie r, R. W. (1979). Prostaglandins in the betaadrenergic and baroreceptor-m ediated secretion of renin. Am J Physiol, 236 (5), F472477. Betker, A. L., Szturm, T., Moussavi, Z. K., & Nett, C. (2006). Video game-based exercises for balance rehabilitation: a single-subject design. Arch Phys Med Rehabil, 87 (8), 11411149. Braith, R. W., & Beck, D. T. (2008). Resistance exercise: training adaptations and developing a safe exercise prescription. Heart Fail Rev, 13 (1), 69-79. Braith, R. W., Schofield, R. S., Hill, J. A., Case y, D. P., & Pierce, G. L. (2008). Exercise training attenuates progressive decline in brachial artery reactivity in heart transplant recipients. J Heart Lung Transplant, 27 (1), 52-59.
120 Braith, R. W., Welsch, M. A., Feigenbaum, M. S., Kluess, H. A., & Pepine, C. J. (1999). Neuroendocrine activation in h eart failure is modified by endurance exerci se training. J Am Coll Cardiol, 34 (4), 1170-1175. Broqvist, M., Dahlstrom, U., Karlsson, E., & Lar sson, J. (1992). Muscle energy metabolism in severe chronic congestive heart failure--effect of treatment with enalapril. Eur Heart J, 13(9), 1217-1224. Cadnapaphornchai, M. A., Gurevich, A. K., Wein berger, H. D., & Schrier, R. W. (2001). Pathophysiology of sodium and water retention in heart failure. Cardiology, 96 (3-4), 122131. Chaudhry, S. I., Wang, Y., Concato, J., Gill, T. M., & Krumholz, H. M. (2007). Patterns of weight change preceding hospitalization for heart failure. Circulation, 116 (14), 15491554. Cicoira, M., Zanolla, L., Franceschini, L., Ro ssi, A., Golia, G., Zamboni, M., et al. (2001). Skeletal muscle mass independently predic ts peak oxygen consumption and ventilatory response during exercise in noncachectic patients with chronic heart failure. J Am Coll Cardiol, 37 (8), 2080-2085. Cider, A., Tygesson, H., Hedberg, M., Seligman, L., Wennerblom, B., & Sunnerhagen, K. S. (1997). Peripheral muscle training in patient s with clinical signs of heart failure. Scand J Rehabil Med, 29 (2), 121-127. Clark, A. L. (2006). Origin of symptoms in chronic heart failure. Heart, 92(1), 12-16. Clark, A. L., Poole-Wilson, P. A., & Coats, A. J. (1996). Exercise limitation in chronic heart failure: central role of the periphery. J Am Coll Cardiol, 28 (5), 1092-1102. Clark, A. L., Volterrani, M., Swan, J. W., & Co ats, A. J. (1997). The increased ventilatory response to exercise in chronic heart failure: relation to pulmonary pathology. Heart, 77(2), 138-146. Cleland, J. G., Louis, A. A., Rigby, A. S., Jan ssens, U., & Balk, A. H. (2005). Noninvasive home telemonitoring for patients with heart failure at high risk of recurrent admission and death: the Trans-European Network-Ho me-Care Management System (TEN-HMS) study. J Am Coll Cardiol, 45 (10), 1654-1664. Coats, A. J., Adamopoulos, S., Radaelli, A., McCa nce, A., Meyer, T. E., Bernardi, L., et al. (1992). Controlled trial of physical trai ning in chronic heart failure. Exercise performance, hemodynamics, ventilation, and autonomic function. Circulation, 85 (6), 2119-2131. Cohen, S. L., Chelland, S., Ball, K. T., & LeMura, L. M. (2002). Effects of fixed ratio schedules of reinforcement on exercise by college students. Percept Mot Skills, 94 (3 Pt 2), 11771186.
121 Conraads, V. M., Beckers, P., Bosmans, J., De Cler ck, L. S., Stevens, W. J., Vrints, C. J., et al. (2002). Combined endurance/resistance trai ning reduces plasma TNF-alpha receptor levels in patients with chronic heart failure and coronary artery disease. Eur Heart J, 23(23), 1854-1860. Conraads, V. M., Beckers, P., Vaes, J., Martin, M., Van Hoof, V., De Maeyer, C., et al. (2004). Combined endurance/resistance training reduc es NT-proBNP levels in patients with chronic heart failure. Eur Heart J, 25 (20), 1797-1805. Conraads, V. M., Vanderheyden, M., Paelinck, B., Ve rstreken, S., Blankoff, I., Miljoen, H., et al. (2007). The effect of endurance training on exercise capacity following cardiac resynchronization therapy in chronic heart failure patients: a pilot trial. Eur J Cardiovasc Prev Rehabil, 14 (1), 99-106. Corra, U., Mezzani, A., Giannuzzi, P., & Tavazz i, L. (2002). Chronic heart failure-related myopathy and exercise training: A devel oping therapy for heart failure symptoms. Prog Cardiovasc Dis, 45(2), 157-172. Cugno, M., Agostoni, P., Mari, D., Meroni, P. L ., Gregorini, L., Bussotti, M., et al. (2005). Impaired bradykinin response to ischaemia and exercise in patients with mild congestive heart failure during angioten sin-converting enzyme treatment. Relationships with endothelial function, coagulation and inflammation. Br J Haematol, 130 (1), 113-120. Davies, L. C., Francis, D., Jurak, P., Kara, T., Pi epoli, M., & Coats, A. J. (1999). Reproducibility of methods for assessing baroreflex sensitiv ity in normal controls and in patients with chronic heart failure. Clin Sci (Lond), 97 (4), 515-522. Davis, P. G., Ferguson, M. A., Alderson, N. L., Pa te, R. R., Bodary, P. F., & Durstine, J. L. (2005). Effect of exercise duration on plasma endothelin-1 concentration. J Sports Med Phys Fitness, 45(3), 419-423. Delagardelle, C., Feiereisen, P., Autier, P., Shita, R., Krecke, R., & Beissel, J. (2002). Strength/endurance training versus enduran ce training in conges tive heart failure. Med Sci Sports Exerc, 34(12), 1868-1872. Delagardelle, C., Feiereisen, P., Krecke, R., Essamr i, B., & Beissel, J. (1999). Objective effects of a 6 months' endurance and strength trai ning program in outpatients with congestive heart failure. Med Sci Sports Exerc, 31 (8), 1102-1107. Dishman, R. K., Sallis, J. F., & Orenstein, D. R. (1985). The determinants of physical activity and exercise. Public Health Rep, 100 (2), 158-171. Doutreleau, S., Piquard, F., Lonsdor fer, E., Rouyer, O., Lampert, E., Mettauer, B., et al. (2004). Improving exercise capacity, 6 wk training tends to reduce circulati ng endothelin after heart transplantation. Clin Transplant, 18 (6), 672-675. Drazner, M. H., & de Lemos, J. A. (2005). Une xpected BNP levels in patients with advanced heart failure: a tale of caution and promise. Am Heart J, 149 (2), 187-189.
122 Drexler, H. (1999). Nitric oxide synthases in the failing human heart: a doubled-edged sword? Circulation, 99 (23), 2972-2975. Drexler, H., Riede, U., Munzel, T., Konig, H., Funke, E., & Just, H. (1992). Alterations of skeletal muscle in chronic heart failure. Circulation, 85 (5), 1751-1759. Dunnigan, A., Staley, N. A., Smith, S. A., Pierpont M. E., Judd, D., Benditt, D. G., et al. (1987). Cardiac and skeletal muscle abnormalitie s in cardiomyopathy: comparison of patients with ventricular tachycardia or congestive heart failure. J Am Coll Cardiol, 10 (3), 608618. Effing, T. W., van Meeteren, N. L., van Asbec k, F. W., & Prevo, A. J. (2006). Body weightsupported treadmill training in chronic incomplete spinal cord injury: a pilot study evaluating functional health status and quality of life. Spinal Cord, 44 (5), 287-296. Ellery, S., Pakrashi, T., Paul, V., & Sack, S. ( 2006). Predicting mortality and rehospitalization in heart failure patients with home mon itoring--the Home CARE pilot study. Clin Res Cardiol, 95 Suppl 3, III29-35. Ennezat, P. V., Malendowicz, S. L., Testa, M., Colo mbo, P. C., Cohen-Solal, A., Evans, T., et al. (2001). Physical training in patients with chronic heart failure enhances the expression of genes encoding antioxidative enzymes. J Am Coll Cardiol, 38 (1), 194-198. Erbs, S., Linke, A., Gielen, S., Fiehn, E., Walther, C., Yu, J., et al. (2003). Exercise training in patients with severe chronic heart failure: impact on left ventricular performance and cardiac size. A retrospective analysis of th e Leipzig Heart Failure Training Trial. Eur J Cardiovasc Prev Rehabil, 10 (5), 336-344. Executive summary: HFSA 2006 Comprehensive Heart Failure Practice Guideline. (2006). J Card Fail, 12(1), 10-38. Feiereisen, P., Delagardelle, C., Vaillant, M., Lasar, Y., & Beissel, J. (2007). Is strength training the more efficient training modality in chronic heart failure? Med Sci Sports Exerc, 39(11), 1910-1917. Feliz, B., Witt, D. R., & Harris, B. T. (2003). Propionic acidemia: a neuropathology case report and review of prior cases. Arch Pathol Lab Med, 127 (8), e325-328. Fonarow, G. C., & Corday, E. (2004). Overview of acutely decompen sated congestive heart failure (ADHF): a report from the ADHERE registry. Heart Fail Rev, 9 (3), 179-185. Francis, G. S. (2007). Emerging strategies to reduce conges tion in decompensated heart failure: New drugs and devices. Paper presented at the Heart Fa ilure Society of America Annual Meeting, Washington, D.C. Fredriksen, B., & Mengshoel, A. M. (2000). The e ffect of static tracti on and orthoses in the treatment of knee contractures in preschool children with j uvenile chronic arthritis: a single-subject design. Arthritis Care Res, 13 (6), 352-359.
123 Froelicher, V. F., & Myers, J. (2006). Exercise and the Heart (5th ed.). Philade lphia: Elsevier. Fuchs, M., & Drexler, H. (2004). Chronic heart failure and proinflammatory cytokines: possible role of physical exercise. Exerc Immunol Rev, 10 56-65. Gademan, M. G., Swenne, C. A., Verwey, H. F., van der Laarse, A., Maan, A. C., van de Vooren, H., et al. (2007). E ffect of exercise training on autonomic derangement and neurohumoral activation in chronic heart failure. J Card Fail, 13 (4), 294-303. Gary, R. (2006). Exercise self-effi cacy in older women with diasto lic heart failure: results of a walking program and education intervention. J Gerontol Nurs, 32 (7), 31-39; quiz 40-31. Gheorghiade, M., Gattis, W. A., O'Connor, C. M., Adams, K. F., Jr., Elkayam, U., Barbagelata, A., et al. (2004). Effects of tolvaptan, a va sopressin antagonist, in patients hospitalized with worsening heart failure: a randomized controlled trial. Jama, 291(16), 1963-1971. Giannuzzi, P., Temporelli, P. L., Corra, U., & Tavazzi, L. (2003). Antiremodeling effect of longterm exercise training in patient s with stable chronic heart fail ure: results of the Exercise in Left Ventricular Dysfunction and Chr onic Heart Failure (ELVD-CHF) Trial. Circulation, 108(5), 554-559. Gielen, S., Adams, V., Linke, A., Erbs, S., Mobius-Winkler, S., Schubert, A., et al. (2005). Exercise training in chronic heart failure: correlation between reduced local inflammation and improved oxidative capacity in the skeletal muscle. Eur J Cardiovasc Prev Rehabil, 12(4), 393-400. Gielen, S., Adams, V., Mobius-Winkler, S., Li nke, A., Erbs, S., Yu, J., et al. (2003). Antiinflammatory effects of exercise training in the skeletal muscle of patients with chronic heart failure. J Am Coll Cardiol, 42 (5), 861-868. Goldberg, L. R., Piette, J. D., Walsh, M. N., Frank, T. A., Jaski, B. E., Smith, A. L., et al. (2003). Randomized trial of a daily electronic home monitoring system in patients with advanced heart failure: the Weight Monitori ng in Heart Failure (WHARF) trial. Am Heart J, 146(4), 705-712. Goldfarb, N., Weston, C., Hartmann, C. W., Sikiri ca, M., Crawford, A., He, H., et al. (2004). Impact of appropriate pharmaceutical thera py for chronic conditions on direct medical costs and workplace productivity: a review of the literature. Dis Manag, 7(1), 61-75. Goldsmith, S. R. (2006). Is there a cardiovascular rationale for the use of combined vasopressin V1a/V2 receptor antagonists? Am J Med, 119 (7 Suppl 1), S93-96. Gordon, A., Tyni-Lenne, R., Jansson, E., Kaijser L., Theodorsson-Norheim, E., & Sylven, C. (1997). Improved ventilation and decreased symp athetic stress in chronic heart failure patients following local enduran ce training with leg muscles. J Card Fail, 3 (1), 3-12. Gross, R. D., Steinhauer, K. M., Zajac, D. J., & Weissler, M. C. (2006). Direct measurement of subglottic air pressure while swallowing. Laryngoscope, 116 (5), 753-761.
124 Guazzi, M., Tumminello, G., Matturri, M., & Guazzi, M. D. (2003). Insulin ameliorates exercise ventilatory efficiency and oxygen uptake in pa tients with heart failure-type 2 diabetes comorbidity. J Am Coll Cardiol, 42 (6), 1044-1050. Gustafsson, T., Bodin, K., Sylven, C., Gordon, A., Tyni-Lenne, R., & Jansson, E. (2001). Increased expression of VEGF following exercise training in patients with heart failure. Eur J Clin Invest, 31 (4), 362-366. Hambrecht, R., Fiehn, E., Weigl, C., Gielen, S ., Hamann, C., Kaiser, R., et al. (1998). Regular physical exercise corrects e ndothelial dysfunction and improve s exercise capacity in patients with chronic heart failure. Circulation, 98 (24), 2709-2715. Hambrecht, R., Fiehn, E., Yu, J., Niebauer, J., We igl, C., Hilbrich, L., et al. (1997). Effects of endurance training on mitochondrial ultrastructure and fiber type distribution in skeletal muscle of patients with stable chronic heart failure. J Am Coll Cardiol, 29 (5), 1067-1073. Hambrecht, R., Gielen, S., Linke A., Fiehn, E., Yu, J., Walther, C., et al. (2000). Effects of exercise training on left vent ricular function and peripheral resistance in patients with chronic heart failure: A randomized trial. Jama, 283(23), 3095-3101. Hambrecht, R., Schulze, P. C., Gi elen, S., Linke, A., Mobius-Winkl er, S., Erbs, S., et al. (2005). Effects of exercise training on insulin-like gr owth factor-I expre ssion in the skeletal muscle of non-cachectic patients with chronic heart failure. Eur J Cardiovasc Prev Rehabil, 12 (4), 401-406. Hambrecht, R., Schulze, P. C., Gielen, S., Linke, A., Mobius-Wi nkler, S., Yu, J., et al. (2002). Reduction of insulin-like growth factor-I expression in the skeletal muscle of noncachectic patients with chronic heart failure. J Am Coll Cardiol, 39 (7), 1175-1181. Hambrecht, R., Wolf, A., Giele n, S., Linke, A., Hofer, J., Erbs, S., et al. (2000). Effect of exercise on coronary endotheli al function in patients with coronary artery disease. N Engl J Med, 342 (7), 454-460. Hammer, A., Nilsagard, Y., Forsberg, A., Pepa, H., Skargren, E., & Oberg, B. (2005). Evaluation of therapeutic riding (Sweden)/hippotherapy (United States). A single-subject experimental design study replicated in eleven patients with multiple sclerosis. Physiother Theory Pract, 21(1), 51-77. Harrington, D., Anker, S. D., Chua, T. P., Webb -Peploe, K. M., Ponikowski, P. P., PooleWilson, P. A., et al. (1997). Skeletal muscle function and its relation to exercise tolerance in chronic heart failure. J Am Coll Cardiol, 30 (7), 1758-1764. Hayashi, K., Sugawara, J., Komine, H., Maeda, S., & Yokoi, T. (2005). Effects of aerobic exercise training on the stiffness of central and peripheral arteries in middle-aged sedentary men. Jpn J Physiol, 55 (4), 235-239.
125 Heidenreich, P. A., Ruggerio, C. M., & Massie B. M. (1999). Effect of a home monitoring system on hospitalization and resource us e for patients with heart failure. Am Heart J, 138(4 Pt 1), 633-640. Hensen, J., Abraham, W. T., Durr, J. A., & Sc hrier, R. W. (1991). Aldosterone in congestive heart failure: analysis of determin ants and role in sodium retention. Am J Nephrol, 11 (6), 441-446. Hornig, B., Maier, V., & Drexler, H. (1996). P hysical training improves endothelial function in patients with chronic heart failure. Circulation, 93 (2), 210-214. Huang, C. N., Chen, C. H., & Chung, H. Y. (2006) Application of facial electromyography in computer mouse access for pe ople with disabilities. Disabil Rehabil, 28 (4), 231-237. Hunt, S. A. (2005). ACC/AHA 2005 guideline update for the di agnosis and management of chronic heart failure in the adult: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Update the 2001 Guidelines for the Evaluation and Management of Heart Failure). J Am Coll Cardiol, 46 (6), e1-82. Indolfi, C., Maione, A., Volpe, M., Rapacciuolo, A., Esposito, G., Ceravolo, R., et al. (1994). Forearm vascular responsive ness to alpha 1and alpha 2-adrenoceptor stimulation in patients with congestive heart failure. Circulation, 90 (1), 17-22. Johansson, P., Dahlstrom, U., & Brostrom, A. (2006). Factors and inte rventions influencing health-related quality of life in patients with heart failure: a review of the literature. Eur J Cardiovasc Nurs, 5 (1), 5-15. Joo, K. C., Brubaker, P. H., MacDougall, A., Sa ikin, A. M., Ross, J. H., & Whaley, M. H. (2004). Exercise prescription using resting he art rate plus 20 or perceived exertion in cardiac rehabilitation. J Cardiopulm Rehabil, 24 (3), 178-184; quiz 185-176. Kannel, W. B., Kannel, C., Paffenbarger, R. S ., Jr., & Cupples, L. A. (1987). Heart rate and cardiovascular mortality: the Framingham Study. Am Heart J, 113 (6), 1489-1494. Kato, M., Kinugawa, T., Omodani, H., Osaki, S ., Ahmmed, G. U., Ogino, K., et al. (1996). Responses of plasma norepinephrine and re nin-angiotensin-aldosterone system to dynamic exercise in patients w ith congestive heart failure. J Card Fail, 2(2), 103-110. Kelly, R. A., & Smith, T. W. (1997). Cy tokines and cardiac contractile function. Circulation, 95(4), 778-781. Keteyian, S. J., Brawner, C. A., Schairer, J. R., Le vine, T. B., Levine, A. B., Rogers, F. J., et al. (1999). Effects of exercise training on chr onotropic incompetence in patients with heart failure. Am Heart J, 138 (2 Pt 1), 233-240.
126 Kiilavuori, K., Naveri, H., Leinonen, H., & Harkonen, M. (1999). The effect of physical training on hormonal status and exertional hormonal re sponse in patients with chronic congestive heart failure. Eur Heart J, 20 (6), 456-464. King, A. C. (2001). Interventions to pr omote physical activity by older adults. J Gerontol A Biol Sci Med Sci, 56 Spec No 2 36-46. Kjaer, A., Appel, J., Hildebrandt, P., & Petersen, C. L. (2004). Basal and exercise-induced neuroendocrine activation in patients with heart failure and in normal subjects. Eur J Heart Fail, 6 (1), 29-39. Kobayashi, N., Tsuruya, Y., Iwasawa, T., Iked a, N., Hashimoto, S., Yasu, T., et al. (2003). Exercise training in patients with chronic heart failure improves endothelial function predominantly in the trained extremities. Circ J, 67 (6), 505-510. Kodiath, M., Kelly, A., & Shively, M. (2005). Impr oving quality of life in patients with heart failure: an innovative behavioral intervention. J Cardiovasc Nurs, 20 (1), 43-48. Konstam, M. A. (2007). Strategies to reduce congestion in ADHF: Lessons learned. Paper presented at the Heart Failure Society of America Annual Meeting, Washington, D.C. Krum, H., & Katz, S. D. (1998). Effect of e ndothelin-1 on exercise-induced vasodilation in normal subjects and in patie nts with heart failure. Am J Cardiol, 81 (3), 355-358. Kubo, S. H., Rector, T. S., Bank, A. J., Williams, R. E., & Heifetz, S. M. (1991). Endotheliumdependent vasodilation is attenuated in patients with heart failure. Circulation, 84 (4), 1589-1596. Lapidos, K. A., Kakkar, R., & McNally, E. M. (2004). The dystrophin glycoprotein complex: signaling strength and integrity for the sarcolemma. Circ Res, 94(8), 1023-1031. Larsen, A. I., Aukrust, P., Aarsland, T., & Dick stein, K. (2001). Effect of aerobic exercise training on plasma levels of tumor necrosis fa ctor alpha in patients with heart failure. Am J Cardiol, 88 (7), 805-808. Larsen, A. I., Gjesdal, K., Hall, C., Aukrust, P., Aarsland, T., & Dickstein, K. (2004). Effect of exercise training in patients with heart failure: a pilot study on autonomic balance assessed by heart rate variability. Eur J Cardiovasc Prev Rehabil, 11(2), 162-167. Lechat, P., Hulot, J. S., Escolano, S., Mallet, A., Leizorovicz, A., WerhlenGrandjean, M., et al. (2001). Heart rate and cardiac rhythm relationships with bi soprolol benefit in chronic heart failure in CIBIS II Trial. Circulation, 103(10), 1428-1433. Lindsay, D. C., Anand, I. S., Bennett, J. G., Pepper, J. R., Yacoub, M. H., Rothery, S. M., et al. (1994). Ultrastructural analysis of skelet al muscle. Microvascular dimensions and basement membrane thickness in chronic heart failure. Eur Heart J, 15 (11), 1470-1476.
127 Lowery, S. L., Massaro, R., & Yancy, C. W., Jr. (2004). Advances in the management of acute and chronic decompensated heart failure. Lippincotts Case Manag, 9 (2 Suppl), S1-15; quiz S16-17. Maeda, S., Otsuki, T., Iemitsu, M., Kamioka, M., Sugawara, J., Kuno, S., et al. (2006). Effects of leg resistance training on ar terial function in older men. Br J Sports Med, 40 (10), 867869. Maguire, S. M., Nugent, A. G., McGurk, C., Johnston, G. D., & Nicholls, D. P. (1998). Abnormal vascular responses in human ch ronic cardiac failure are both endothelium dependent and endothelium independent. Heart, 80 (2), 141-145. Maiorana, A., O'Driscoll, G., Cheetham, C., Co llis, J., Goodman, C., Ra nkin, S., et al. (2000). Combined aerobic and resistance exercise training improves functional capacity and strength in CHF. J Appl Physiol, 88 (5), 1565-1570. Mancini, D. M., Coyle, E., Coggan, A., Beltz, J., Ferraro, N., Montai n, S., et al. (1989). Contribution of intrinsic skeletal muscle ch anges to 31P NMR skeletal muscle metabolic abnormalities in patients with chronic heart failure. Circulation, 80 (5), 1338-1346. Mancini, D. M., Walter, G., Reichek, N., Lenkins ki, R., McCully, K. K., Mullen, J. L., et al. (1992). Contribution of skeletal muscle at rophy to exercise intolerance and altered muscle metabolism in heart failure. Circulation, 85 (4), 1364-1373. Massie, B., Conway, M., Yonge, R., Frostick, S ., Ledingham, J., Sleight P., et al. (1987). Skeletal muscle metabolism in patients with congestive heart failure: relation to clinical severity and blood flow. Circulation, 76 (5), 1009-1019. Massie, B. M., Conway, M., Rajagopalan, B., Yonge, R., Frostick, S., Ledingham, J., et al. (1988). Skeletal muscle metabolism during exercise under ischemic conditions in congestive heart failure. Evidence for abnormalities unrelated to blood flow. Circulation, 78(2), 320-326. McCleary, L. (2002). Equipoise in clinical nursing research. Can J Nurs Res, 34 (3), 49-60. McHorney, C. A., Ware, J. E., Jr., & Raczek, A. E. (1993). The MOS 36-Item Short-Form Health Survey (SF-36): II. Psychometric and clinical tests of validity in measuring physical and mental health constructs. Med Care, 31(3), 247-263. Mettauer, B., Zoll, J., Sanchez, H., Lampert, E., Ribera, F., Veksler, V., et al. (2001). Oxidative capacity of skeletal muscle in heart failure patients versus sedentary or active control subjects. J Am Coll Cardiol, 38 (4), 947-954. Meyer, K., & Laederach-Hofmann, K. (2003). Eff ects of a comprehensive rehabilitation program on quality of life in patients with chronic heart failure. Prog Cardiovasc Nurs, 18(4), 169-176.
128 Michaud, T. C., & Nawoczenski, D. A. (2006). Th e influence of two diffe rent types of foot orthoses on first metatarsophalangeal joint kinematics during gait in a single subject. J Manipulative Physiol Ther, 29 (1), 60-65. Minotti, J. R., Pillay, P., Oka, R., Wells, L., Christoph, I., & Massie, B. M. (1993). Skeletal muscle size: relationship to muscle function in heart failure. J Appl Physiol, 75 (1), 373381. Mitchell, J. H., & Victor, R. G. (1996). Neural control of th e cardiovascular system: insights from muscle sympathetic ne rve recordings in humans. Med Sci Sports Exerc, 28 (10 Suppl), S60-69. Morey, C. E., Cilo, M., Berry, J., & Cusick, C. (2003). The effect of Aricept in persons with persistent memory disorder following traumatic brain injury: a pilot study. Brain Inj, 17(9), 809-815. Nakamura, M. (1999). Peripheral vascular remodeli ng in chronic heart failu re: clinical relevance and new conceptualization of its mechanisms. J Card Fail, 5 (2), 127-138. Nakamura, M., Arakawa, N., Yoshida, H., Maki ta, S., Niinuma, H., & Hiramori, K. (1998). Vasodilatory effects of B-type natriuretic pe ptide are impaired in patients with chronic heart failure. Am Heart J, 135 (3), 414-420. Nakamura, M., Chiba, M., Ueshima, K., Arakawa, N., Yoshida, H., Makita, S., et al. (1995). Impaired cholinergic periphe ral vasodilation and its rela tionship to hyperemic calf blood flow response and exercise intolerance in patients with chronic heart failure. Int J Cardiol, 48(2), 139-146. Nakamura, M., Yoshida, H., Arakawa, N., Miz unuma, Y., Makita, S., & Hiramori, K. (1996). Endothelium-dependent vasodilatation is not selectively impaired in patients with chronic heart failure secondary to valvular heart disease and congenital heart disease. Eur Heart J, 17(12), 1875-1881. Nakamura, M., Yoshida, H., Makita, S., Arakaw a, N., Niinuma, H., & Hiramori, K. (1997). Potent and long-lasting vasodilatory effect s of adrenomedullin in humans. Comparisons between normal subjects and patien ts with chronic heart failure. Circulation, 95 (5), 12141221. Niebauer, J., Clark, A. L., WebbPeploe, K. M., Boger, R., & Coats, A. J. (2005). Home-based exercise training modulates prooxidant substrates in patients with chronic heart failure. Eur J Heart Fail, 7 (2), 183-188. Niebauer, J., Clark, A. L., WebbPeploe, K. M., & Coats, A. J. (2005). Exercise training in chronic heart failure: effects on pro-inflammatory markers. Eur J Heart Fail, 7 (2), 189193.
129 Niebauer, J., Pflaum, C. D., Clark, A. L., Strasbur ger, C. J., Hooper, J., Poole-Wilson, P. A., et al. (1998). Deficient insulin-like growth factor I in chronic h eart failure predicts altered body composition, anabolic deficiency, cy tokine and neurohormonal activation. J Am Coll Cardiol, 32 (2), 393-397. Nieminen, M. S., Bohm, M., Cowie, M. R., Drexler, H., Filippatos, G. S., Jondeau, G., et al. (2005). Executive summary of the guidelines on the diagnosis and treatment of acute heart failure: the Task Force on Acute Heart Failure of the European Society of Cardiology. Eur Heart J, 26 (4), 384-416. Nishiyama, Y., Minohara, M., Ohe, M., Hirai, Y., Katoh, A., Miyamoto, T., et al. (2006). Effect of physical training on insulin resistance in patients with chronic heart failure. Circ J, 70(7), 864-867. O'Leary, C. J., & Jones, P. W. (2000). The le ft ventricular dysfunction questionnaire (LVD-36): reliability, validity, and responsiveness. Heart, 83 (6), 634-640. Oka, R. K., DeMarco, T., & Haskell, W. L. ( 2005). Effect of treadmill testing and exercise training on self-efficacy in patients with heart failure. Eur J Cardiovasc Nurs, 4 (3), 215219. Okamoto, T., Masuhara, M., & Ikuta, K. (2008) Relationship between plasma endothelin-1 concentration and cardiovascular responses dur ing high-intensity eccentric and concentric exercise. Clin Physiol Funct Imaging, 28 (1), 43-48. Otsuka, Y., Takaki, H., Okano, Y., Satoh, T., Aihara, N., Matsumoto, T., et al. (2003). Exercise training without ventricular remodeling in patients with moderate to severe left ventricular dysfunction early afte r acute myocardial infarction. Int J Cardiol, 87 (2-3), 237-244. Passino, C., Severino, S., Poletti, R., Piepoli, M. F., Mammini, C., Cler ico, A., et al. (2006). Aerobic training decreases B-type natriu retic peptide expression and adrenergic activation in patients with heart failure. J Am Coll Cardiol, 47 (9), 1835-1839. Piepoli, M., Adamopoulos, S., Bern ardi, L., Sleight, P., & Coats, A. J. (1995). Sympathetic stimulations by exercise-stress testing and by dobutamine infusion induce similar changes in heart rate variability in patients with chronic heart failure. Clin Sci (Lond), 89(2), 155164. Piepoli, M., Clark, A. L., Volterrani, M., Adam opoulos, S., Sleight, P., & Coats, A. J. (1996). Contribution of muscle afferents to th e hemodynamic, autonomic, and ventilatory responses to exercise in patie nts with chronic heart failure: effects of phy sical training. Circulation, 93 (5), 940-952. Piepoli, M., Ponikowski, P., Clark, A. L., Bana siak, W., Capucci, A., & Coats, A. J. (1999). A neural link to explain the "m uscle hypothesis" of exercise intolerance in chronic heart failure. Am Heart J, 137 (6), 1050-1056.
130 Piepoli, M. F., Kaczmarek, A., Francis, D. P., Davies, L. C., Rauchhaus, M., Jankowska, E. A., et al. (2006). Reduced periphera l skeletal muscle mass and abnormal reflex physiology in chronic heart failure. Circulation, 114 (2), 126-134. Piepoli, M. F., Ponikowski, P. P., Volterrani, M. Francis, D., & Coats, A. J. (1999). Aetiology and pathophysiological implications of oscillatory ventilation at rest and during exercise in chronic heart failure. Do Cheyne and St okes have an important message for modernday patients with heart failure? Eur Heart J, 20 (13), 946-953. Quittan, M., Sturm, B., Wiesinger, G. F., Pacher, R., & Fialka-Moser, V. (1999). Quality of life in patients with chronic heart failure: a ra ndomized controlled trial of changes induced by a regular exercise program. Scand J Rehabil Med, 31 (4), 223-228. Ramanaiah, N. V., Franzen, M., & Schill, T. (198 3). A psychometric study of the State-Trait Anxiety Inventory. J Pers Assess, 47 (5), 531-535. Rector, T. S., Kubo, S. H., & Cohn, J. N. (1993). Validity of the Minnesota Living with Heart Failure questionnaire as a measure of ther apeutic response to enalapril or placebo. Am J Cardiol, 71(12), 1106-1107. Reinlib, L., & Abraham, W. (2003). Recovery from heart failure with circulatory assist: a working group of the National, Heart, Lung, and Blood Institute. J Card Fail, 9 (6), 459463. Resnick, B., Magaziner, J., Orwig, D., & Zimm erman, S. (2002). Evaluating the components of the Exercise Plus Program: rati onale, theory and implementation. Health Educ Res, 17(5), 648-658. Rich, M. W., & Nease, R. F. (1999). Cost-effectiven ess analysis in clinical practi ce: the case of heart failure. Arch Intern Med, 159(15), 1690-1700. Roth, A., Kajiloti, I., Elkayam, I., Sander, J., Kehati, M., & Golovner, M. (2004). Telecardiology for patients with chronic heart failure: the 'SHL' experience in Israel. Int J Cardiol, 97 (1), 49-55. Roveda, F., Middlekauff, H. R., Rondon, M. U., Reis, S. F., Souza, M., Nast ari, L., et al. (2003). The effects of exercise training on sympathe tic neural activation in advanced heart failure: a randomized controlled trial. J Am Coll Cardiol, 42 (5), 854-860. Schaufelberger, M., Eriksson, B. O., Grimby, G. Held, P., & Swedberg, K. (1997). Skeletal muscle alterations in patients with chronic heart failure. Eur Heart J, 18 (6), 971-980. Scheier, M. F., Matthews, K. A., Owens, J. F., Magovern, G. J., Sr., Lefebvre, R. C., Abbott, R. A., et al. (1989). Dispositional optimism and recovery from coronary artery bypass surgery: the beneficial effects on physical and psychological well-being. J Pers Soc Psychol, 57(6), 1024-1040.
131 Schrier, R. W. (2006). Water and sodium retenti on in edematous disorders: role of vasopressin and aldosterone. Am J Med, 119 (7 Suppl 1), S47-53. Schrier, R. W., & Abraham, W. T. (1999). Hormones and hemodynamics in heart failure. N Engl J Med, 341 (8), 577-585. Schrier, R. W., Cadnapaphornchai, M. A., & Umenishi, F. (2001). Water-losing and waterretaining states: role of water channels and vasopressin receptor antagonists. Heart Dis, 3(3), 210-214. Schrier, R. W., & Fassett, R. G. (1998). Pathogene sis of sodium and water retention in cardiac failure. Ren Fail, 20(6), 773-781. Schrier, R. W., Ohara, M., Rogachev, B., X u, L., & Knotek, M. (1998). Aquaporin-2 water channels and vasopressin antagoni sts in edematous disorders. Mol Genet Metab, 65 (4), 255-263. Selektor, Y., & Weber, K. T. (2008). The salt-a vid state of congestive heart failure revisited. Am J Med Sci, 335 (3), 209-218. Selig, S. E., Carey, M. F., Menzies, D. G., Patte rson, J., Geerling, R. H., Williams, A. D., et al. (2004). Moderate-intensity resistance exerci se training in patient s with chronic heart failure improves strength, endurance, heart rate variability, and forearm blood flow. J Card Fail, 10(1), 21-30. Senden, P. J., Sabelis, L. W., Zonderland, M. L., Hulzebos, E. H., Bol, E., & Mosterd, W. L. (2005). The effect of physica l training on workload, upper leg muscle function and muscle areas in patients w ith chronic heart failure. Int J Cardiol, 100 (2), 293-300. Smart, N., & Marwick, T. H. (2004). Exercise training for patients w ith heart failure: a systematic review of factors that improve mortality and morbidity. Am J Med, 116 (10), 693-706. Smith, B., Forkner, E., Zaslow, B., Krasuski, R. A., Stajduhar, K., Kwa n, M., et al. (2005). Disease management produces limited quality-of-life improvements in patients with congestive heart failure: evidence from a randomized trial in community-dwelling patients. Am J Manag Care, 11 (11), 701-713. Stanley, R. (1998). Applying the four prin ciples of ethics to continence care. Br J Nurs, 7 (1), 4451. Stevenson, L. W., Steimle, A. E., Fonarow, G., Kermani, M., Kermani, D., Hamilton, M. A., et al. (1995). Improvement in exercise capacity of candidates awaiting heart transplantation. J Am Coll Cardiol, 25 (1), 163-170. Stoykov, M. E., Stojakovich, M., & Stevens, J. A. (2005). Beneficial effects of postural intervention on prehensile action for an indi vidual with ataxia resulting from brainstem stroke. NeuroRehabilitation, 20 (2), 85-89.
132 Sturm, B., Quittan, M., Wiesinger, G. F., Stan ek, B., Frey, B., & Pacher, R. (1999). Moderateintensity exercise training with elements of step aerobics in patients with severe chronic heart failure. Arch Phys Med Rehabil, 80(7), 746-750. Sullivan, M. J., Green, H. J., & Cobb, F. R. (1990) Skeletal muscle biochemistry and histology in ambulatory patients with long-term heart failure. Circulation, 81 (2), 518-527. Sullivan, M. J., Higginbotham, M. B., & Cobb, F. R. (1988). Increased exercise ventilation in patients with chronic heart fa ilure: intact ventilatory co ntrol despite hemodynamic and pulmonary abnormalities. Circulation, 77 (3), 552-559. Swan, J. W., Anker, S. D., Wa lton, C., Godsland, I. F., Clark, A. L., Leyva, F., et al. (1997). Insulin resistance in chronic heart failure: relation to severity and etiology of heart failure. J Am Coll Cardiol, 30 (2), 527-532. Swedberg, K., Cleland, J., Dargie, H., Drexler, H., Follath, F., Komajda, M., et al. (2005). Guidelines for the diagnosis and treatment of chronic heart failure: executive summary (update 2005): The Task Force for the Diagnos is and Treatment of Chronic Heart Failure of the European Society of Cardiology. Eur Heart J, 26 (11), 1115-1140. Tousoulis, D., Charakida, M., & Stefanadis, C. (2005). Inflammation and endothelial dysfunction as therapeutic targets in patients with heart failure. Int J Cardiol, 100 (3), 347-353. Toyo-Oka, T., Kawada, T., Nakata, J., Xie, H., Ur abe, M., Masui, F., et al. (2004). Translocation and cleavage of myocardial dystrophin as a co mmon pathway to advanced heart failure: a scheme for the progression of cardiac dysfunction. Proc Natl Acad Sci U S A, 101 (19), 7381-7385. Tyni-Lenne, R., Gordon, A., Jensen-Urstad, M., De ncker, K., Jansson, E., & Sylven, C. (1999). Aerobic training involving a minor muscle mass shows greater efficiency than training involving a major muscle mass in chronic heart failure patients. J Card Fail, 5 (4), 300307. Umstattd, M. R., & Hallam, J. (2007). Older a dults' exercise behavior : roles of selected constructs of social-cognitive theory. J Aging Phys Act, 15 (2), 206-218. van Tol, B. A., Huijsmans, R. J., Kroon, D. W ., Schothorst, M., & Kwakkel, G. (2006). Effects of exercise training on cardiac performance, exercise capacity and quality of life in patients with heart failure: A meta-analysis. Eur J Heart Fail Verbalis, J. G. (2006). Whole-body volume re gulation and escape from antidiuresis. Am J Med, 119(7 Suppl 1), S21-29. Verbalis, J. G., Murase, T., Ecel barger, C. A., Nielsen, S., & Kne pper, M. A. (1998). Studies of renal aquaporin-2 expression during renal escap e from vasopressin-induced antidiuresis. Adv Exp Med Biol, 449 395-406.
133 Vescovo, G., Serafini, F., Dalla Libera, L., Leprot ti, C., Facchin, L., Tenderini, P., et al. (1998). Skeletal muscle myosin heavy chains in hear t failure: correlation between magnitude of the isozyme shift, exercise capacit y, and gas exchange measurements. Am Heart J, 135(1), 130-137. Vescovo, G., Volterrani, M., Zennaro, R., Sandri, M., Ceconi, C., Lorusso, R., et al. (2000). Apoptosis in the skeletal muscle of patients with heart failur e: investigation of clinical and biochemical changes. Heart, 84 (4), 431-437. Volaklis, K. A., & Tokmakidis, S. P. (2005). Resi stance exercise training in patients with heart failure. Sports Med, 35 (12), 1085-1103. Volterrani, M., Clark, A. L., Ludman, P. F., Swan, J. W., Adamopoulos, S ., Piepoli, M., et al. (1994). Predictors of exercise cap acity in chronic heart failure. Eur Heart J, 15 (6), 801809. Wang, X. Y., Masilamani, S., Niel sen, J., Kwon, T. H., Brooks, H. L., Nielsen, S., et al. (2001). The renal thiazide-sensitive Na-Cl cotranspor ter as mediator of the aldosterone-escape phenomenon. J Clin Invest, 108 (2), 215-222. Webel, A. R., Frazier, S. K., Moser, D. K., & Le nnie, T. A. (2007). Daily variability in dyspnea, edema and body weight in heart failure patients. Eur J Cardiovasc Nurs, 6 (1), 60-65. Weiner, B. (1985). An attributional theory of achievement mo tivation and emotion. Psychol Rev, 92(4), 548-573. Weissman, M. M., Sholomskas, D., Pottenger, M ., Prusoff, B. A., & Locke, B. Z. (1977). Assessing depressive symptoms in five psychiatric populations: a validation study. Am J Epidemiol, 106(3), 203-214. Wielenga, R. P., Erdman, R. A., Huisveld, I. A., Bo l, E., Dunselman, P. H., Baselier, M. R., et al. (1998). Effect of exercise training on quality of life in patients with chronic heart failure. J Psychosom Res, 45 (5), 459-464. Willenheimer, R., Rydberg, E., Cline, C., Broms, K., Hillberger, B., Oberg, L., et al. (2001). Effects on quality of life, symp toms and daily activity 6 mont hs after termination of an exercise training programme in heart failure patients. Int J Cardiol, 77 (1), 25-31. Wilson, J. R., Ferraro, N., & Wiener, D. H. (1985). Effect of the sympathetic nervous system on limb circulation and metabolism during exercise in patients with heart failure. Circulation, 72 (1), 72-81. Wilson, J. R., Fink, L., Maris, J., Ferraro, N., Power-Vanwart, J., Eleff, S., et al. (1985). Evaluation of energy metabolism in skeletal mu scle of patients with heart failure with gated phosphorus-31 nuclea r magnetic resonance. Circulation, 71 (1), 57-62. Witte, K. K., & Clark, A. L. (2005). Cycle exer cise causes a lower ventilatory response to exercise in chronic heart failure. Heart, 91 (2), 225-226.
134 Wong, L. L., & Verbalis, J. G. (2002). Systemic diseases associated with disorders of water homeostasis. Endocrinol Metab Clin North Am, 31 (1), 121-140. Wray, D. W., Nishiyama, S. K., Donato, A. J., Sander, M., Wagner, P. D., & Richardson, R. S. (2007). Endothelin-1-mediated vasoconstricti on at rest and during dynamic exercise in healthy humans. Am J Physiol Heart Circ Physiol, 293 (4), H2550-2556. Yancy, C. W. (2007). Defining decompensated heart failure : What is known and what is not known. Paper presented at the Heart Failure Society of America Annual Meeting, Washington, DC. Yancy, C. W., Lopatin, M., Stevenson, L. W., De Marco, T., & Fonarow, G. C. (2006). Clinical presentation, management, and in-hospital outcomes of patients admitted with acute decompensated heart failure with preserved systolic function: a report from the Acute Decompensated Heart Failure Natio nal Registry (ADHERE) Database. J Am Coll Cardiol, 47 (1), 76-84. Yoshida, H., Nakamura, M., Akatsu, T., Arakawa, N., & Hiramori, K. (1998). Effects of nitric oxide inhibition on basal for earm blood flow in patients with nonischemic chronic heart failure. Heart Vessels, 13 (3), 142-146. Yoshizumi, M., Perrella, M. A., Burnett, J. C., Jr ., & Lee, M. E. (1993). Tumor necrosis factor downregulates an endothelial nitric oxide s ynthase mRNA by shortening its half-life. Circ Res, 73 (1), 205-209.
135 BIOGRAPHICAL SKETCH Andrea Boyd was born in Anchorage, Alaska. She received her bachelors degree in psychology with a m inor in the life sciences, graduating cum laude, from Cl emson University in 1993. After receiving her degree she continue d her gymnastics coaching career. She then moved to Athens, Georgia and began to pursue he r masters degree. She graduated in 1998 with a masters degree in exercise science with her thesis titled, Evaluation of body image in premenarcheal, competitive, womens artistic gymn asts and age-, height-, and weight-matched controls. Her mentors while at the University of Georgia were Drs. Patrick OConnor, Rod Dishman, Kirk Cureton, and Richard Lewis. She worked as a research assistant on an American Heart Association grant funded research project under Dr. Di shman investigating exercise training and its effects on the bl ood pressure of African-Americans. After graduating from the University of Geor gia, Andrea continued her coaching career. In 2003, she chose to leave elite coaching after over 15 years and began a new career in nursing. She enrolled in the first class of the Univers ity of Floridas Accelerated BSN program. In 2004, she received her BSN and progressed into the BSN-PhD program. In 2005, she received her MSN as a Clinical Nurse Specialist in Adult Medi cal/Surgical Health. Her area of specialty is cardiology with a focus in heart failure. While pursuing her degrees her at UF Andreas mentors have been many. The faculty that have mentor ed Andrea have been Drs. James Jessup, Eileen Handberg, Laura Sutton, Jennifer Elder, David Criswell, Meredeth Rowe, Mary Jo Snider, Joyce Stechmiller, Keith Mueller, and Shawn Kneip. The outside individuals that mentored Andrea are Dr. Peggy Guinn, Arlene Davis, Debbie Irish, Dr Connie Uphold, Dr. Ron Shorr, Dr. Christy Carter, Dr. Richard Schofield, and the many nurses who taught her how to be a nurse. Andrea was awarded a UF Alumni Graduate Fellowship and then a VA Pre-Doctoral Fellowship. These awards, not only an honor, a llowed the financial freedom for Andrea to
136 devote full time to school and MANY unique oppor tunities not ordinarily available to doctoral students. She was able to have unique indepe ndent studies with individuals knowledgeable in her area of study, to travel to conferences for her area of specialty, a nd to work on research grants otherwise not available to her. For thes e opportunities she is eter nally grateful to the funding bodies of both of these organizat ions for providing such opportunities. Andrea was able to present a poster at the Southern Nursing Research Society in 2008, Innovative Methodology for Nursing Research. This same poster received first place at the 2007 University of Florida, College of Nursing Research Day and Malasanos Lectureship. In addition, she has presented three other posters to this Research Day in 2006 and 2007. She also did an oral presentation in 2008 for the same c onference. In 2007, she also presented at the Institute of Aging and GRECC Research Sy mposium on her research area. In 2008, she presented a poster for the VA Research Day. Andrea will graduate in August of 2008 with her PhD in Nursing Science and a minor in Exercise Physiology. She plans to publish her work from her dissertation a nd to collaborate with the original studys PI to publish additional findi ngs from the original study. Additionally, she plans to continue her work in the areas of the physiological benefits of exercise and the merger of home-based exercise programs and the tele-hea lth/remote monitoring fields of health care as a nurse researcher and educator.