The purpose of this dissertation was to examine the associations among social support, gender, relationship quality, and self-care behaviors in patients with heart failure (HF). Specific aims were to: 1) compare the psychometric properties of the 12-item and the 9-item European Heart Failure Self-Care Behavior Scale (EHFScBS) when used to measure self-care behaviors in patients with HF in the United States; 2) determine whether gender moderated the association between perceived social support and daily sodium intake in patients with HF; and 3) determine whether a patient’s gender and relationship with the primary family caregiver (spousal or non-spousal) moderated the association between relationship quality and HF self-care. For the first specific aim, reliability and validity of the 12-item and the 9-item EHFScBS were compared by measures of internal consistency of reliability, item analysis, exploratory factor analysis, and hypothesis testing. The 12-item and the 9-item EHFScBS were valid and reliable when used to measure self-care behaviors in the United States. The 9-item EHFScBS was more homogenous with dimensions closer to the originally proposed dimensions when compared to the 12-item EHFScBS. The second specific aim was addressed by secondary analysis of data from The RICH Heart Program HF Registry. A hierarchical multiple regression model was used to determine whether gender moderated the association between social support and daily sodium intake. In our study, gender did not moderate the association between social support and daily sodium intake. The third specific aim was addressed using hierarchical multiple regression analyses to determine whether gender or relationship type moderated the association between relationship quality and HF self-care. Baseline data from an ongoing longitudinal, randomized controlled trial of a patient-caregiver dyadic intervention program were used in the analysis. Relationship quality was measured using the patient version of the Dyadic Relationship Scale, which consists of two subscales: positive dyadic interaction and negative dyadic strain. Heart failure self-care was measured using the 9-item EHFScBS. Positive dyadic interaction was associated with better HF self-care in female patients or patients with a non-spousal relationship with their family caregiver. This dissertation has fulfilled important gaps in the evidence base for the self-care of patients with HF. The findings from this dissertation provided evidence for the validity and reliability of the 9-item EHFScBS when used to measure HF self-care behaviors. It also provided recommendations for future research to measure directed social support to adopt a specific behavior, such as eating low sodium diet, instead of measuring the social support in general. It also emphasized the importance of examining the quality of the relationship between patients and their family caregivers. The findings also pointed to the need for interventions targeting the relationship quality to be tailored according to the patients’ gender or their relationship type with their family caregiver (spousal or non-spousal). A better understanding of how social factors can impact self-care behaviors in patients with HF is essential, as it gives researchers and healthcare providers the prospect of developing effective interventions to improve HF self-care.

Alarming rates of suicide among production farmers have prompted researchers to investigate factors associated with depressive symptoms among this population. Aspects of farm life and farming can contribute to higher levels of depressive symptoms. Higher levels of depression can also increase an individual’s risk of injury and development of chronic disease, impacting overall quality of life. Despite the approximate 3.5 million farm women in the U.S., current research has focused on the male farmer. Men and women have different responses to stressors, and women in general have a higher prevalence of depressive symptoms. Farm women can be further subjected to stressors associated with farming as an occupation and their gendered role within the agrarian culture. The large number of farm women affected, the relationship of chronic depressive symptoms on health and quality of life, the lack of current research available, and the rising rates of suicide and depressive symptoms among farmers emphasize the need for further investigation of farm women and depressive symptoms. The overall purpose of this dissertation was to 1) explore the current state of the science of farm women and depressive symptoms and identify variables commonly associated with depressive symptoms among farm women, 2) identify variables influencing levels of depressive symptoms within farm women aged 50 and over and identify differences between those women with high depressive symptoms and those with low depressive symptoms, and 3) establish the reliability and validity of the 12-item John Henry Active Coping Scale (JHAC-12) within the sample. A systematic review of the literature revealed that there is a need for more research with strong study designs regarding farm women and depressive symptoms within the context of their environment, culture, and occupation. The review identified multidimensional factors from farm women’s lives that influence their level of depressive symptoms. Farm women’s ethnicity, the agrarian culture, family and social relations, as well as specific demographics were identified as key variables associated with an increased risk of higher depressive symptoms. Because of the identification of the multi-dimensional factors, the use of the Modified Biopsychosocial Model (MBPS) was selected as a framework for continued research as it depicts the interrelationship between the factors and their influence on farm women’s depressive symptoms. The MBPS was applied to data from 358 farm women aged 50 and older from a larger cohort study, and a secondary analysis was performed. Multivariable binary logistic regression was used to identify those variables associated with depressive symptoms among farm women. Depressive symptoms were predicted by race/ethnicity, years of education, adequacy of income for vacation, perceived health status, perceived stress score, and active coping score. Significant differences between those farm women with low CES-D score (< 16) and those with high CES-D score (≥ 16) were noted. Race/ethnicity, years of education, adequate income for vacation and retirement, reported health status of fair or better, perceived stress score, active coping score and satisfaction from farm work were all significant between groups. Women who were non-White, had less education, reported income not adequate for vacation or retirement, reported poor health, higher levels of perceived stress, lower levels of active coping and who were not satisfied with farm work were more likely to be in the high CES-D group. A principal component analysis with direct oblimin rotation in a sample population of older farm women (n=458) identified two dominant themes of the JHAC-12: “commitment to hard work” and “self-efficacy.” The instrument component structure reflects the culture of the agrarian society. In the two-component solution, 2 items were removed from the scale after revealing low values of communality (< .3). The item reduction resulted in more refined scale, increasing explained variance by 4.1% with less items. Cronbach’s of the JHAC-12 (α = .78) and JHAC-10 (α = .76) indicated high levels of reliability for both scales. Rotation of the items resulted in a simple structure with high loadings within items, no major-cross-loadings and little correlation between components (r = .29), supporting both convergent and discriminant validity in this population. The ability of the JHAC to encompass the socio-culture aspects of active coping among farm women and obtain a quantifiable result supports the JHAC as an important tool to utilize in future studies of depressive symptoms and farm women with use of the JHAC-10 in future studies of farm women decreasing the burden of the participants. Although there are limitations within each document, each section adds to the science of farm women and depression symptoms and provides directions for future research. The major gaps identified were: 1) the need for current research with stronger study designs, 2) studies of farm women across their life spans, 3) the need for focused studies among minority and migrant women, 4) an understanding of farm women and their leisure time, and 5) a broader application of the MBPS theory to include a large number of social variables shown to be associated with farm women and depressive symptoms that were not available in the dataset.

The purpose of this dissertation was to explore the physical and psychological responses to the combination of major trauma (Injury Severity Score [ISS] > 15) and variable periods of immobility. Specific aims were to: 1) develop a conceptual model that illustrates physiological and psychological alterations that occur after injury and subsequent immobility, and their association with skeletal muscle responses and recovery; 2) evaluate daily measures of skeletal muscle strength (bicep and quadricep) using dynamometry and skeletal muscle (rectus femoris and biceps brachii) muscle thickness measured with ultrasound in patients after major trauma; and 3) assess the predictive ability of anxiety and depressive symptoms after traumatic injury on delayed ambulation (> 48 hours) following hospital admission. Specific Aim 1 was addressed by development of a conceptual model to describe the association between injury responses, immobility and skeletal muscle after trauma based on a comprehensive review of the state of the science. This model guided the research reported in Aims 2 and 3. The second specific aim was addressed with the conduct of an observational study in which we evaluated daily skeletal muscle strength with dynamometry and muscle thickness with ultrasound to evaluate the impact of trauma and immobility on skeletal muscle in patients after major trauma (n = 19). Participants with delayed ambulation after trauma (more than 48 hours immobility) demonstrated significantly less muscle strength compared with those who had early ambulation (bicep: delayed ambulation 12.9 ± 3.8, early ambulation 17.7 ± 4.7, p = 0.004; quadriceps: delayed ambulation 9.9 ± 3.1, early ambulation 17.1 ± 4.6, p = 0.001). Muscle thickness was unchanged over time in those with delayed ambulation; however, in those who ambulated early, muscle thickness significantly increased by 0.17 cm (p = 0.008) from baseline to day 5. The third specific aim was addressed with data collected during the same observational study of patients after trauma (n = 19). Participants provided measures of anxiety and depressive symptoms at baseline. Anxiety was not a predictor of delayed ambulation; however, depressive symptoms increased the likelihood of delayed ambulation by 67% (Odds Ratio [OR]: 1.67, 95% CI: 1.02 – 2.72, p = 0.041). Early ambulation was associated with significantly greater muscle strength and thickness as determined by dynamometry and muscle ultrasound, and depressive symptoms significantly increased the likelihood of delayed ambulation. Systematic evaluation of the association between trauma injury, immobility, skeletal muscle function and structure, and psychological state will provide an opportunity for the appropriate evaluation after injury and development of effective, tailored interventions to improve short- and long-term physiological and psychological recovery.

Ischemic stroke is the leading cause of disability world-wide and affects over 800,000 people per year in the United States. The majority of these strokes are ischemic due to a blockage of blood flow to the brain. Damage to the brain occurs at the onset of stroke, neuronal cell death is irreversible and therefore, quick treatment to remove blockage is critical factor in the recovery from stroke. Mechanical thrombectomy as a treatment for ischemic stroke provides an ideal opportunity to collect blood distal and proximal to the cerebral thrombus to examine neurochemical changes occurring during stroke. The purpose of this dissertation was to explore the trajectory of neurochemical changes that occur in response to ischemic stroke during the first 72 hours and the physiological response from stroke patients to improve stroke outcomes. The specific aims were to: 1) to determine whether venous blood gases predict infarct volume and/or mortality in acute ischemic stroke in young male rats; 2) determine whether venous blood gases predict infarct and edema volume, and/or mortality in acute ischemic stroke in aged male and female rats; 3) compare the presence and relative concentrations of acid/base and electrolytes in static blood distal to thrombus and in peripheral blood drawn from adults who received thrombectomy for ischemic stroke and identify associations to postreperfusion functional outcomes. Specific Aim One was addressed by evaluation of young (three-month old) Sprague-Dawley rats that underwent permanent or transient middle cerebral artery occlusion (MCAO). Pre- and post-MCAO venous samples from permanent and transient models provided pH, carbon dioxide, oxygen, bicarbonate, glucose, hematocrit, hematocrit, and electrolyte values of ionized calcium, potassium and sodium. The analyses indicated that mean differences in the blood gas and electrolytes between pre- to post-MCAO and pH and iCa2+ were predictors of infarct volume in the permanent MCAO model. The second aim was addressed by evaluation of aged (18 month old) male and female rats pre-MCAO, post-MCAO, and at 72 hours of permanent MCAO venous blood gas samples (pH, carbon dioxide, oxygen, bicarbonate, glucose, hematocrit, hematocrit, and electrolyte concentrations of ionized calcium, potassium and sodium). Changes in pH (from pre-MCAO to post-MACO and post-MCAO to 72 hours) and changes in Na+ and iCa2+ (from post-MCAO to 72 hours) were predictors of infarct volume and edema volume, respectively in both sexes. Cox regression revealed there was a 3.25 times increased risk for mortality based on changes (cut-off range within -2.00 to - 7.00) in bicarbonate levels (pre- to post-MCAO). The third aim was addressed by evaluation of acid/base balance (pH, carbon dioxide, oxygen, bicarbonate, ionized calcium, potassium and sodium) of ischemic stroke patients who underwent mechanical thrombectomy. Our results suggests sex differences matter in ischemic stroke populations. Significant differences occur within proximal blood between the sexes. Additionally, females had approximately 2.5 hour increased time between stroke symptom onset to thrombectomy completion time (described as infarct time). Changes in bicarbonate and base deficit were predictors of infarct time, but only in our female population.

Heart failure (HF) is one of the top causes of mortality in the United States and globally. In order to combat the high mortality rates of this disease, medical technology, including internal cardioverter defibrillators (ICD) and left ventricular assist devices (LVAD), have become one of the most common treatments. Over the past 10 years the utilization of these cardiac devices has increased exponentially, which has created a new phenomenon of how we discuss death with patients who have one of these devices. The purpose of this dissertation is to increase understanding of the end-of-life decision making processes and current experiences that patients with a cardiac device are having. This dissertation includes four original manuscripts that focus on patients with a cardiac device and their experiences with decision-making at the end-of-life. The first paper is a data-based paper that examines experiences of patients with an ICD and what factors are associated with having a conversation with their providers about end-of-life. The second paper is an integrative review of the literature regarding what is currently known about end-of-life with an LVAD. The third paper is a psychometric evaluation of the Control Attitudes Scale-Revised (CAS-R) for patients with an LVAD. The fourth paper is a data-based manuscript that looks at patients with an LVAD and their attitudes and experiences with end-of-life conversations with providers and next-of-kin and the impact of cognition on these attitudes and experiences. The findings of this dissertation will hopefully inform providers of patients with cardiac devices about their patients end-of-life decision making processes. It will also demonstrate the gaps that are currently in practice, and ideally be able expand on how to assist patients and providers on improving communication about end-of-life decision making.

Individuals with type 2 diabetes mellitus (T2DM) are at risk for developing life-threatening comorbidities such as cardiovascular disease (CVD). As a consequence, T2DM is associated with increased morbidity and mortality and decreased quality of life, thus highlighting the importance of prevention of T2DM. Further, the prevalence of T2DM is substantially greater in rural populations compared to urban populations, making rural individuals particularly appropriate targets for T2DM prevention. T2DM is a largely preventable disease that is associated with modifiable risk factors such as poor diet, sedentary lifestyle, and obesity. Lifestyle interventions to improve these modifiable risk factors have been used to decrease the risk of developing T2DM. There is little evidence that supports lifestyle interventions as a means to decrease T2DM risk in rural populations with prediabetes, the precursor of T2DM. The purpose of this dissertation was to determine whether rural-living individuals with prediabetes would improve modifiable risk factors, specifically diet quality by following a lifestyle intervention; thereby, decreasing their risk of developing T2DM. Specific aims for this dissertation were to, 1) examine and synthesize data from dietary interventions used to reduce risk of T2DM in rural populations on order to identify gaps and guide future research, 2) critically evaluate validity and reliability of indices used to determine diet quality in research, and 3) determine the effect of a risk reduction program on improving diet quality and glucose control (as a measure of T2DM risk) in rural adults with prediabetes and CVD risk factors. Specific aim one was achieved by a review and synthesis of literature focused on lifestyle and dietary interventions used in rural populations to decrease the risk of developing T2DM. Common goals in these studies were a decrease in weight, decrease in dietary fat and calories, and an increase in physical activity. Decreased weight and increased physical activity were demonstrated in all eight studies, and a decrease in T2DM incidence was also demonstrated in one of the studies. However, diet quality was not adequately assessed in the majority of the studies. Furthermore, none of the studies were randomized controlled trials and only half used a control group. It was concluded that research using a more robust design is needed to determine the effect of lifestyle changes, specifically diet, on T2DM risk in rural populations. Specific aim two was addressed by a critical analysis of six common indices of dietary quality. Validity and reliability of the Healthy Eating Index, the Alternative Healthy Eating Index, the DASH diet score, the Diet Quality Index-Revised, the Healthy Diet Indicator, and the Diet Quality Score were examined. Five of the six indices are valid and reliable tools for measure diet quality but all five rely on an extensive food frequency questionnaire that may be burdensome for participants. The Diet Quality Score does not provide adequate evidence to support its use in research. It was concluded that a short, reliable, and validated diet screener may be useful in research. Specific aim three was addressed by a secondary data analysis of a longitudinal, randomized controlled study of rural residents with CVD risk factors and prediabetes. Diet quality, measured by the Mediterranean Diet Adherence Screener (MEDAS), and glucose control, measured by hemoglobin A1c, were analyzed in a subpopulation of 62 participants with prediabetes. Neither diet quality nor glucose control improved between baseline, four month, and 12 month post intervention. The reliability and validity of the MEDAS in this population is not known and may have been a factor in the lack of intervention effect related to diet quality. Participants were also not informed of their prediabetes status, thus it is not known if this knowledge would have made an impact on the outcomes of the study. In addition, the small sample size limits the statistical power to determine changes between the intervention and control groups. It was concluded that further research is needed to determine if a high quality diet will reduce T2DM risk in this rural population Considering the disproportionate prevalence of T2DM in rural populations compared to their urban counterparts, the results of this dissertation demonstrate a continued need for interventions that decrease modifiable risk factors associated with this disease. Interventions that target obesity, poor diet quality, and sedentary lifestyles in at-risk rural populations that are culturally tailored are needed to decrease risk of developing T2DM and the comorbidities associated with this preventable disease.

The purpose of this dissertation was to examine depression, experiences of work-related racism, and occupational stress among black nurses. Nursing is a highly stressful and demanding profession that can negatively affect health. Underscoring this is the high rate of depression experienced by nurses. In fact, nurses experience depression at a rate twice that of individuals in other occupations. Examining depression in nurses can provide insights that can inform measures addressing the psychological health of this group. This may be particularly important in black nurses who, in addition to the already high occupational stress associated with nursing, may experience additional stress due to experiences of racism in the work environment. To better understand these factors, the specific aims of this dissertation were to: (1) evaluate the current state of the science of depression in registered nurses; (2) examine the psychometric properties of the two racism on the job subscales of the Perceived Racism Scale in black registered nurses; and (3) evaluate whether past-year or lifetime experiences of work-related racism and occupational stress predicted depressive symptoms and whether, controlling for depressive symptoms, past-year and lifetime experiences of work-related racism predicted occupational stress in a cohort of black registered nurses. For specific aim one a systematic review of the literature on depression in nurses was conducted. This review highlighted factors that underlie the high rates of depression among nurses, and the individual as well as work-related variables that contribute to nurses’ susceptibility to depression. For specific aim two the psychometric properties of two subscales of the Perceived Racism Scale in a sample of black registered nurses were evaluated. The two subscales were past year experiences of racism on the job (ROTJ-Y) and lifetime experiences of racism on the job (ROTJ-L). Reliability for each of the subscales was assessed by examining internal consistency. Construct validity was examined using principal components analysis to evaluate the factor structure of each subscale and by testing the hypothesis that job-related racism is predictive of workplace stress. These analyses demonstrated that the ROTJ-Y and ROTJ-L are valid and reliable instruments for the measurement of yearly and lifetime experiences of racism on the job in black registered nurses. Specific aim three was addressed by examining whether past-year or lifetime experiences of racism on the job and occupational stress were predictive of depression and whether work-related racism predicted occupational stress in a sample of black nurses. Multiple linear regression analyses were conducted to examine if (1) either past-year or lifetime experiences of work-related racism, and occupational stress predicted depression and (2) either past-year or lifetime experiences of racism predicted occupational stress, with control variables depressive symptoms, years of experience as a registered nurse, primary nursing practice position, work setting, work shift, and work status. Results indicated that experiences of work-related racism and occupational stress were not significant predictors of depression but that both past-year and lifetime experiences of racism were significant predictors of occupational stress. The results of the research conducted for this dissertation highlight the effects of depression on nurses as well as the relationship between race-based discrimination at work and occupational stress among black registered nurses. This evidence can inform the development of future strategies to improve the well-being of nurses in the workplace in general and especially of black nurses.

The purpose of this dissertation was to evaluate the association of psychological distress with cardiac disease, events, and mortality. Specific aims were to: 1) to evaluate the association between hostility level and recurrence of acute coronary syndrome (ACS) and all-cause mortality in patients with coronary heart disease (CHD); 2) to evaluate the psychometric properties of the Brief Symptom Inventory (BSI) hostility and anxiety subscales in a group of incarcerated participants at high risk of cardiovascular disease; and 3) to evaluate the association of patient and caregiver psychological state with quality of life in both patient and caregiver, and postoperative complications after cardiac surgery. Specific aim one was addressed through a secondary analysis of data collected during the Patient Response to Myocardial Infarction following a Teaching Intervention Offered by Nurses trial to determine whether hostility was a predictor of ACS recurrence and mortality. Hostility was common after ACS and predicted all-cause mortality. Hostility did not predict recurrent ACS. Specific aim 2 was addressed in a secondary analysis of baseline data from a randomized controlled trial in male prisoners. Participants completed the BSI at baseline prior to the intervention. Internal consistency reliability was good for both subscales (Cronbach’s alpha - hostility 0.83, anxiety 0.81). Items from the two dimensions were analyzed together using exploratory factor analysis with varimax rotation. Two dimensions, anxiety and hostility, were identified. Construct validity was supported; those with high anxiety and hostility reported a greater number of days where their self-reported health was rated as fair or poor. Those prisoners with less perceived control had higher levels of anxiety and hostility. Specific aim 3 was addressed through a prospective, descriptive correlational study that measured patient and caregiver anxiety, hostility and depressive symptoms, at baseline to determine whether these predicted quality of life using a multilevel dyadic analysis; and to evaluate the association of baseline anxiety, hostility and depressive symptoms and quality of life with postoperative complications and mortality. Anxiety, hostility, and depressive symptoms were common in both cardiac patients and their caregiver. Psychological state influenced quality of life in both dyad members, but was not associated with complications.

Heart failure is a progressive condition that affects over 5.7 million Americans and costs associated with heart failure account for 2-3 % of the national health care budget. The high rates of morbidity and mortality along with increased costs from readmissions associated with advanced heart failure have led to the exploration of advanced treatments such as left ventricular assist devices (LVADs). LVADS have demonstrated morbidity and mortality benefit but cost remains extensive with costs per quality-adjusted years > $400,000. With this in mind, it is important to identify those who are most likely to benefit from an LVAD to avoid unfavorable outcomes and cost. Although general guidelines and criteria for patient eligibility have been established, choosing patients for LVAD implantation remains challenging. A new focus on patient selection involves the presence of frailty. While frailty has been studied in the elderly population and in patients undergoing cardiac surgery, frailty in patients undergoing left ventricular assist device (LVAD) remains controversial. The purpose of this dissertation was to examine measures of frailty in patients undergoing LVAD implantation. The specific aims of this dissertation were to: (1) identify a feasible frailty measure in adults with end-stage heart failure who underwent LVAD implantation by testing the hypothesis that frailty would predict 30 day rehospitalization rates using Fried’s criteria, Short Physical Performance Battery test, handgrip strength, serum albumin and six minute walk test (2) Determine whether frailty measures improve 3 months post LVAD implantation (3) compare sensitivity of these three measures to change in frailty. Surgical approaches, including heart transplantation and LVAD implantation, for patients with end-stage heart failure was discussed in this dissertation. Data from two subsets of participants who underwent LVADS at the University of Kentucky between 2014 and 2017 were included in the analysis for this dissertation. In the first study, we found that none of the measures are good predictors of frailty in patients with advanced heart failure who undergo LVAD implantation. Handgrip was the only marker of frailty that predicted 30 day readmission but the relationship was a negative association. In the second study, six-minute walk and low serum albumin levels reflect short-term improvement in frailty. These simple measures may be used to determine those patients who are responsive to LVAD implantation. The findings of these studies filled some gaps in our understanding of markers of frailty in patients undergoing LVADs. We gained a better understanding of which markers of frailty are likely to improve in most people after LVAD implantation and thus frailty should not preclude candidate selection for an LVAD. Subsequently, more research is needed to investigate these markers and outcomes.

Background: Heart failure (HF) is a leading cause of hospitalization, readmissions, and death in the United States. Patients hospitalized for HF are at risk for readmission, in- hospital mortality, and early post-discharge death. In the United States, inpatient care has been estimated to cost $83,980 over the lifetime of each patient with HF. The majority of patients with HF depend on caregiver support for successful HF self-care, which is essential for optimal patient outcomes. Support from caregivers is thought to be important for better self-care, and lower readmission and mortality rates. Yet, there are few studies considering the influence of caregivers on HF patient self-care, readmission, and mortality. Objective: The purpose of my dissertation was to determine the influence of HF caregivers on patient self-care, readmission, and mortality. The specific aims of this dissertation were to: (1) to determine if caregiver depressive symptoms mediate the relationship between family functioning and caregiver quality of life, (2) to determine if there is an association between living arrangements (living with someone vs. living alone) and all-cause readmission and death in patients with HF, and (3) to determine the efficacy of an in-hospital, multi-session, educational intervention for caregivers on heart failure patients’ self-care and 30 day readmission rate, and to evaluate the efficacy of the intervention on caregivers’ knowledge, self-efficacy and perceived control. Methods: Specific aim one was addressed by a secondary analysis of data from one- hundred and forty-three HF caregivers recruited from an outpatient clinic. Multiple regression with mediation analysis was used to determine whether depressive symptoms mediated the relationship between family functioning as measured using the three scales of the Family Assessment Device (i.e., general, problem-solving, communication) and caregiver quality of life. Specific aim two was addressed by a retrospective chart review of all 398 patients with a primary diagnosis of HF admitted to an academic medical center in one year. We collected data on patient sociodemographic, clinical characteristics, and patient living condition. The independent association of living alone with all-cause readmission or all-cause death was evaluated using Cox proportional hazards modeling adjusting for covariates. Specific aim three was addressed using a two-group (educational intervention for caregivers of patients with heart failure vs. usual educational care), prospective, repeated measures randomized controlled trial of 37 patient and caregiver dyads in which caregivers only received in-hospital HF education. Outcome measures included patient self-care, and patient all-cause readmission or all- cause death, as well as caregiver self-efficacy, knowledge, and perceived control. Patient self-care, and caregiver self-efficacy, knowledge, and perceived control were assessed at baseline (in hospital), at discharge, 7 and 30-days after patient discharge. Patient readmissions and death were assessed by a phone call at 30-days follow-up. The intervention directed only at caregivers consisted of three in-hospital, educational sessions with telephone follow-up. The educational sessions were designed to deliver HF information and skills to caregivers, thereby providing them with the resources needed to improve their self-efficacy, perceived control and HF knowledge thus improving patient self-care and readmission rates. Results: Specific aim one: The three subscales of the Family Assessment Device predicted depressive symptoms (p < 0.001) and caregiver quality of life (p < 0.001). Depressive symptoms also predicting caregiver quality of life (p < 0.001). The inclusion of depressive symptoms in the final model with each subscale of the Family Assessment Device (i.e., general family functioning, problem-solving, communication) decreased the significance of family functioning as a predictor of caregiver quality of life indicating mediation by depressive symptoms. Specific aim two: Heart failure patients living with someone experienced a significantly longer time to rehospitalization than those living alone (290 vs. 201 days, p=0.005). In a Cox regression hazard regression model, adjusting for covariates, patients who lived alone were 1.42 times more likely to be rehospitalized one year after discharge than those who lived with someone (p=0.013). The relationship between living alone and all-cause death was not significant after adjustment for covariates. Specific aim three: A linear mixed-model analysis revealed that patients whose caregiver was in the intervention group had significantly better self- care maintenance (p < 0.001) and self-care management (p < 0.001) across time. Cox survival analysis demonstrated that patients whose caregiver did not receive the educational intervention were 11 times more likely (p=0.002) to experience cardiac readmission than patients whose caregiver did receive the educational intervention. Caregivers who received the educational intervention had higher perceived control (p < 0.001) for up to 30-days post-intervention versus the control group, however, there were no differences between caregiver groups in self-efficacy and HF knowledge. Conclusion: In this dissertation, we found caregivers to play an important part in improving patient outcomes of self-care and readmission after discharge from a hospitalization for HF. Future large-scale studies are needed to develop and test interventions focused on caregivers to improve both patient and caregiver outcomes. Such studies will assist clinicians in understanding how better to support caregivers in their ability to positively influence HF self-care and readmission rates in patients with HF.