| dc.description.abstracteng | BACKGROUND: Chronic heart failure is associated with high morbidity and mortality and impairs all dimensions of HRQOL. Heart failure with preserved ejection fraction (HFpEF) is a complex syndrome in which patients with normal or near-normal systolic LV-function show symptoms and signs of heart failure. Although numerous mechanisms are involved in the pathogenesis of HFpEF, diastolic LV-dysfunction is assumed a central role. HFpEF is primarily a disease of the elderly and accounts for approximately 50% of the cases of chronic heart failure. A number of risk factors for the development of HFpEF are known. These include: age, arterial hypertension, atrial fibrillation, female gender, and metabolic syndrome. There is currently no effective drug therapy for HFpEF. In addition to comprehensive generic and disease-specific HRQOL-Instruments with numerous items, a single health self-assessment question (self-rated health, SRH) provides a simple and comprehensive measurement of HRQOL. SRH is an established predictor of mortality. The role of HRQOL in patients with risk factors for HFpEF development has been poorly understood.
METHODS AND RESULTS: This thesis is based on data from the DIAST-CHF study, an observation study that included a total of 1937 patients with at least one risk factor for the development of HFpEF. The relationship between SRH and the SF-36 subscales were investigated. Significant trends in age, BMI, 6-minute walking distance, LVEF and NT-proBNP plasma levels were found across the SRH groups. Common comorbidities (heart failure, CHD, atrial fibrillation, COPD, PAD, arterial hypertension, hyperlipidemia, diabetes mellitus, hyperuricemia and depression) showed a significant trend to poor SRH. No significant trends to poor SRH have been shown in myocardial infarction, cerebrovascular disease and a history of CABG. Poor SRH showed a highly significant trend towards lower scores in all SF-36 subscales. Unadjusted and age-adjusted, univariate ordinal logistic regression analyzes were carried out for selected predictors and the odds ratios for reporting of a worse SRH were determined. Significant predictors were: Heart failure, COPD, depression, arterial hypertension, PAD, atrial fibrillation, CHD, systolic LV-dysfunction, diabetes mellitus, hyperuricemia, NT-proBNP ≥ 220 ng/L, cerebrovascular disease and E/e' ≥ 15. No significant predictor of poor SRH was a history of CABG. Age-adjusted, E/e' ≥ 15 was no longer a significant predictor of a poorer SRH. The remaining predictors retained their significance in age-adjusted terms. A multivariate ordinal logistic regression analysis for selected predictors of a poor SRH was conducted. The chosen confounding variables were age, female gender, depression, COPD, CHD, diabetes mellitus, atrial fibrillation, systolic LV-dysfunction, and an E/e' ratio ≥ 15. Except for the echocardiographic parameters, all predictors of a poorer SRH were significant. To examine the relationship between HRQOL and outcome, survival analyses with Kaplan-Meier-Plots and Cox-Regression were performed. In the Kaplan-Meier analyses the quartiles of the SF-36 subscales "Physical Component Score" (PCS) and "Physical Functioning" a well as the SRH-Groups showed significant differences in event free survival for the occurrence of a combined endpoint (death from any cause or cardiovascular hospitalization). No significant difference in event free survival was found for the SF-36 Subscale "Mental Component Score". Cox regression analyzes determined the hazard ratios (HR) for all-cause mortality and for the combined endpoint (death or cardiovascular hospitalization) for SRH and the quartiles of PCS. The first (worst) and second PCS-Quartiles were found having significantly higher hazard ratios for the occurrence of death and the combined endpoint than the fourth (best) PCS-Quartile. SRH "Good", "Fair" or "Poor" showed significantly higher hazard ratios for occurrence of death or combined endpoint than patients with SRH "Excellent" or "Very Good".
Prognosis models using binomial logistic regression were constructed for prognosis of death or cardiovascular hospitalization. A regression model was identified as a predictive model that included demographic factors (age, sex), echocardiographic parameters (systolic left ventricular dysfunction and E / e'), and comorbidities (COPD, CHD, atrial fibrillation, arterial hypertension, diabetes mellitus) and NT-proBNP. Overall, this basic forecasting model was found to be significant. Additional regression models were created in which the basic forecasting model was extended with the HRQOL-Instruments SRH and the SF-36 subscales PCS and "Physical Functioning". The SF-36 subscales PCS and "Physical Functioning" were significant predictors in the respective models. Although SRH had a tendency for significance as a predictor, it failed to do so. The extension of the regression models with HRQOL instruments showed an improvement of the prognosis models. Improvement in specificity became clear; however, the improvement in a comparison of the areas under the ROC curves did not reach the level of significance. CONCLUSION: HRQOL tools appear to be promising predictors of risk stratification tools for patients with HFpEF risk factors, and the easy-to-evaluate HRQOL instruments used here have the potential to provide a meaningful contribution to risk stratification. | de |