Association of pulse pressure and aortic root diameter in elderly Chinese patients with chronic heart failure

Background High pulse pressure (PP) and aortic root diameter (AoD) are hallmarks of arterial stiffness or vascular aging and they are considered as risk factors for age-related cardiovascular disease, including heart failure (HF). However, the relationship between PP and AoD in patients with heart failure (HF) is uncertain. This study aimed to evaluate the relationship between PP and AoD in the middle-aged and the elderly with HF. Methods A total of 1,027 Chinese middle-aged and elderly patients with HF, including HF with reduced ejection fraction (HFrEF), HF with mid-range EF (HFmrEF), and HF with preserved EF (HFpEF) were included in this study. Pearson correlation analysis was used to evaluate the relationship between PP and AoD in the three types of HF. Multiple linear regression analysis was performed to assess the factors that affected AoD. Multivariate logistic regression was performed to determine the association between the PP level quartiles and AoD. The results were validated in an independent dataset included a total of 378 consecutive patients with HFrEF hospitalized at the Pingtan Branch of Fujian Medical University Union Hospital (Fujian, China). Results There was a positive correlation between PP and AoD in the middle-aged and the elderly with HFrEF. Multiple linear regression analysis revealed that PP, age, and body mass index (BMI) were independently correlated with AoD in HFrEF patients. In multivariate logistic regression analysis, an increased risk of aortic root dilation was observed in the highest quartile of the PP level compared with the lowest quartile. Age significantly interacted with PP (p = 0.047). A significant association between PP levels and AoD was only observed in patients ≥ 65 years old, but not in patients < 65 years old. In the validation dataset, PP was independently related to AoD in patients with HFrEF (β = 0.205, p = 0.001). Conclusions PP level was independently and positively associated with AoD, especially in the elderly with HFrEF, but not in patients with HFmrEF and HFpEF. Arterial stiffening or vascular aging may play a certain role in the elderly HFrEF patients.


Introduction
Millions of adults worldwide suffer from heart failure (HF), which is associated with higher mortality rates, morbidity, and healthcare costs in the world (1,2).Clinically, three types of HF are recognized: HF with reduced ejection fraction (HFrEF), HF with mid-range ejection fraction (HFmrEF), and HF with preserved ejection fraction (HFpEF) (3).Previous studies demonstrated that pulse pressure (PP) could be used to predict left ventricular hypertrophy and cardiovascular events (4-9).Wei et al. found that high PP predicted all-cause death, nonfatal myocardial infarction, nonfatal stroke, or hospitalization in patients with HF (10).
Several factors influence aortic root dilation, including age, gender, height, weight, body surface area (BSA), body mass index (BMI), and diseases such as hypertension, valvular heart disease, congenital heart disease, cardiomyopathy and ischemic cardiomyopathy (11).Recently, we and others have found that aortic root dilation also is one of the hall-mark of vascular aging or arterial stiffness (12,13) which can lead to isolated systolic hypertension (14)(15)(16) and result in left ventricular remodeling, dysfunction, and heart failure (17,18).A community-based cohort of Afro-American study found that aortic root diameter (AoD) was associated with an increased risk of cardiovascular events (19).The Framingham Heart Study indicated that the risk of incident HF increased with greater AoD at baseline and an increase in AoD over 8 years (20).AoD may be useful as a predictor of cardiovascular events even in the absence of aneurysmatic alterasions (21).
The precise relationship between PP and AoD is debatable (22-25) and there has yet to be a study looking into the association between them in the three types of HF.Furthermore, the relationship between PP and AoD in the middle-aged and the elderly with HF remains to be elucidated.Therefore, we aimed to investigate the relationship between PP and AoD in patients with HF in the present study.

Patients
We investigated the medical records of 1027 consecutive HF patients (age ≥ 45 years old) were hospitalized at the Fujian Medical University Union Hospital (Fujian, China) between January 2015 and December 2018.The patients' medical histories and relevant clinical characteristics were recorded.Patients were excluded from this study if they presented with any of the following conditions: acute myocardial infarction, acute myocarditis, significant valvular heart disease, congenital heart disease, renal failure [estimated glomerular filtration rate (eGFR) < 60 ml/min/1.73m 2 ], malignancy, and chronic inflammatory disease.
The validation dataset included a total of 378 consecutive patients with HFrEF hospitalized at the Pingtan Branch of Fujian Medical University Union Hospital (Fujian, China) between January 2015 and December 2018.The inclusion criteria included the HFrEF patients who underwent Echocardiography.The exclusion criteria included patients lacking echocardiography test, patients ≤ 45 years of age, and other confounding conditions, such as acute myocardial infarction, acute myocarditis, significant valvular heart disease, congenital heart disease, renal failure [estimated glomerular filtration rate (eGFR) < 60 ml/min/1.73m 2 ], malignancy, and chronic inflammatory disease.

Ethics statement
The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013) and approved by the Medical Faculty of Fujian Medical University Union Hospital Ethics Committee (No. 2019KY004).
between the leading edge to the leading edge (L-L) convention of anterior and posterior aortic root wall at the maximal level of the sinuses of Valsalva, as recommended by the American Society of Echocardiography guidelines (28).Left ventricular dimensions were measured based on the American Society of Echocardiography recommendations.Aortic root dilation was defined as an AoD of ≥34 mm and ≥30 mm in males and females, respectively.
Patients were classified into non-smokers and current smokers (continuously smoking one or more cigarettes a day for at least six months) based on their smoking status (29).Hypertension was diagnosed as SBP ≥ 140 mmHg and/or DBP ≥ 90 mmHg and/or taking antihypertensive medications according to the European Society of Cardiology (ESC) guidelines (30).

Definition and type of HF
The diagnosis of HF was based on the symptoms or signs, electrocardiograms (ECG), chest radiographs, and echocardiography (31).HF was categorized as HFpEF if the left ventricular ejection fraction (LVEF) was ≥50%, HFmrEF if LVEF was 40%-49%, and HFrEF if LVEF was <40%.Chronic kidney disease (CKD) was diagnosed when eGFR was <90 ml/ min/1.73m 2 .Coronary heart disease (CHD) was defined as a history of coronary stent implantation or coronary artery bypass graft and myocardial ischemia on ECG, with symptoms typical of angina or myocardial infarction.Stroke was defined as the presence of a definite history of stroke or signs of cerebral infarction on computed tomography or magnetic resonance imaging.

Statistical analysis
The normality of the data was assessed by the Kolmogorov-Smirnov test.Normally distributed variables are presented as mean ± standard deviation and compared via student t test.Non-normally distributed variables are expressed as the median and interquartile range (IQR) and analyzed by Mann-Whitney U test.Categorical variables are expressed as numbers and percentages (%) and compared using the χ 2 test or Fisher's exact test (if theoretical frequency T < 5).1-way ANOVA followed by the Bonferroni post test for multiple comparisons.Linear correlations between variables were assessed using Pearson's correlation analysis.Non-linear associations between variables were assessed using restricted cubic spline analysis.Multiple linear regression analyses were performed to assess the independent determinants of AoD/BSA.The association between AoD/BSA and clinical parameters including patients' age, BMI, blood test indicators, hemodynamic parameters and echocardiographic indicators, was analyzed based on stepwise linear regression.This analysis was conducted by considering AoD/BSA as independent variables, and relevant clinical characteristics were set as dependent variables.Using the logistic regression analysis, PP was grouped by quartiles when analyzing the relationship between PP and aortic dilation risk using logistic regression analysis.The interactions between PP and aortic dilation risk were assessed by introducing a crossproduct term into the regression models.Statistical significance was set at p < 0.05.All data were analyzed using the SPSS (version 19.0, IL, USA) software.

Correlation between PP and AoD in patients with HFrEF
PP and AoD were found to increase with age in patients with HF (Figure 1, Supplementary Figure S1).Therefore, we further explored the association between PP and AoD in patients with HFrEF, HFmrEF, and HFpEF.The results showed that PP was positively associated with AoD in the middle-aged and the elderly with HFrEF (r = 0.380, p < 0.01), but not in patients with HFmrEF (r = 0, p = 0.99) or HFpEF (r = 0.01, p = 0.9) (Figure 2).
The independent association of AoD/BSA with other variables was assessed using multiple linear regression analysis.We observed that PP was independently associated with AoD/ BSA (β = 0.170, p < 0.004), adjusting for age, BMI, SBP, DBP, HR, hemoglobin level, triglyceride (TG), total cholesterol(TC), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), eGFR, LVEF, left ventricular end diastolic diameter (LVED.The independent determinants of AoD/BSA in the fully adjusted models included PP, age, BMI, and ALB (Table 2).
To assess the non-linear association between PP and AoD, a restricted cubic spline analysis was performed.The results indicated that there was no non-linear association between PP and AoD in patients with HFrEF (Figure 3).
Multiple regression analysis was employed to assess the association between PP levels and aortic root dilation in three models.In the univariate logistic regression analysis, with Q1 set as the reference, the PP levels in Q4 were associated with an increased odds ratio (OR) for aortic root dilation [OR = 5.685 95% confidence interval (CI): 1.941-16.646,p for trend = 0.015].After adjusting for age, sex, BSA, smoking, hypertension, we observed that the PP levels in Q4 were associated with an increased OR for aortic root dilation (OR = 4.897, 95% CI: 1.629-14.717,p for trend = 0.043) as compared with the PP levels in Q1.Moreover, when adjusting for complications and hematological markers simultaneously, those in Q4 had a higher risk of aortic root dilation than those in Q1 (Q4: 6.612, 95% CI: 1.838-23.783vs. Q1: 2.411, 95% CI: 1.037-5.607;p for trend = 0.049) (Table 4).Subsequently, stratified analysis was performed to explore further the association between PP and AoD in different population settings, including age, BMI, hypertension, LDL-C, and smoking.An interaction test was performed to assess any significant dependence of the effect modifier on this association.We observed a significant modification of the association between PP levels and AoD (p for interaction = 0.047).A significant association between PP levels and AoD was only observed in patients ≥ 65 years old (OR = 1.753, 95% CI: 1.159-2.652,p for trend = 0.008), but not in patients < 65 years old (p for trend = 0.930).No significant interaction was observed in BMI, LDL-C, smoking status, and hypertension (p for interaction = 0.540, 0.571, 0.470, and 0.289, respectively), indicating that the magnitude of the relationship was the same for different population settings concerning these variables (Table 5).

Validation dataset for relationship of PP and AoD
Based on the discovery data, we validated the relationship between PP and AoD in another dataset.The information for the validation data is shown in Supplementary Table S1.Then, we further validated the association between PP and AoD in patients with HFrEF.The results showed that PP was positively associated with AoD in patients with HFrEF (r = 0.15, p < 0.01; Figure 2D).The correlations were further assessed by linear regression analysis.We observed that PP was independently associated with AoD (β = 0.205, p < 0.001), adjusting for age, BMI, blood test indicators, hemodynamic parameters and echocardiographic indicators (Supplementary Table S2).

Discussion
In our study, age was lowest in HFrEF group compared to the other kinds of heart failure.Although the management of HFrEF has seen significant scientific breakthrough in recent decades, HFrEF is a major public health concern with substantial morbidity and mortality.Disease morbidity and mortality remain high, with a 5-year survival rate of 25% after hospitalization for HFrEF (32).Among the three groups of heart failure types, the LVED of HFrEF was larger than that of the other two groups, suggesting that HErEF was more likely to undergo ventricular remodeling.Deleterious LV remodeling, including increases in  Frontiers in Cardiovascular Medicine end-diastolic and end-systolic volume and reduction in LVEF, is pathognomonic of HFrEF, and the extent of adverse remodeling correlates with risk of hospitalization and death.LV volumes and contractility worsen progressively over time (33).This also suggests that early intervention should be initiated for the treatment of HFrEF.We found that the average level of PP was 44.8 ± 14.0 mmHg in patients with HFrEF, 52.6 ± 16.5 mmHg in patients with HFmrEF, and 60.3 ± 17.5 mmHg in patients with HFpEF in our study.These findings are similar to those of a previous study showing the lowest average level of PP in the HFrEF group (34).This phenomenon may be related to the wide use of angiotensin converting enzyme inhibitors/angiotensin receptor blockers, beta-blockers, and spironolactone in patients with HFrEF, which are known to lower SBP and PP (35).Meanwhile, we found that the systolic blood pressure in the HFpEF group was the highest and their heart rate was the lowest compared to other groups.In contrast, the aortic root diameter was not significantly different between HF classification.Studies (36,37) showed that the lower SBP being a dose-dependent marker of impaired left ventricular contractility.Advanced heart failure is usually associated with low systolic blood pressure (SBP).Studies (38, 39) have been reported that increased resting heart rate is a risk factor for cardiovascular mortality in cardiovascular diseases.In heart failure patients, increased heart  The aorta has a significant influence on left ventricular afterload and hemodynamics.Age-related enlargement of AoD is a characteristic of arterial stiffness that can influence cardiovascular disease (15,22,23).PP, an indicator of pulsatile flow ejected by the heart, is mainly influenced by LVEF and arterial stiffness (40).It has been reported that high PP is usually related to increased SBP and systolic hypertension accelerates arterial stiffness, thereby resulting in CVD (4,(41)(42)(43).Epidemiologic studies have shown that PP is predictive of the incidence of HF in the elderly population (44).Several epidemiological studies have examined the relationship between the proximal aorta and major adverse cardiovascular events.For example, the Cardiovascular Health Study (CHS) (45) and the study by Cuspidi et al. (46) showed that increased AoD was associated with a higher incidence of cardiovascular events.However, the relationship between PP and AoD remains controversial.Cuspidi et al. (24) found a positive association between PP and AoD (r = 0.10, p = 0.004) in a total of 3,366 treated and untreated patients with essential hypertension.(47).In contrast, there was an independent inverse relationship between AoD and PP in the Framingham cohort study (23) and in a study by Agmon et al. (25) The authors considered that the inverse association between AoD and PP might be due to reduced wave reflection as a result of higher aortic compliance in those with larger AoD and considered invasive aortic root PP data as a for possible explanation for this negative association (48).In this study, Pearson correlation analysis showed that PP was positively correlated with AoD in patients with HFrEF.Furthermore, the probability of increased aortic dilation was associated with higher PP quartiles, and a higher risk of aortic dilation was observed in patients with PP ≥ 65 mmHg, which is inconsistent with previously reported results.A possible reason for this discrepancy in the findings between this study and those of the Framingham study is related to the different populations examined.This study focused on the middle-aged and the elderly with HF.In addition, we observed a significant interaction between PP levels and age, indicating that PP and age may have an additive effect on AoD.It is well known that aging leads to arterial stiffness, which makes the arterial wall more susceptible to the harmful effects of PP (49).This phenomenon may explain the positive association of PP levels with AoD observed only in patients ≥ 65 years with HFrEF.
To our knowledge, the present study is the first to show that an increased PP is significantly associated with AoD in patients with HFrEF, but not in patients with HFmrEF or HFpEF.One reason might be that AoD can be influenced by age, height, weight, left ventricular structure (15), and lower the levels of LVED in patients with HFmrEF or HFpEF than that in patients with HFrEF.Appropriate management may be an effective strategy for lowering the risk of aortic dilation in elderly patients with HFrEF.There were no specific interventions focused on PP currently.Therefore, blood pressure management may be beneficial in reducing the risk of aortic dilation in the elderly with HFrEF.To further verify the reliability of our conclusions, we collected the dataset in another hospital to find consistent results from different hospital populations.Finally, the results improved the conclusion reliability.
Certain limitations of this study should be noted.First, this was a limited cross-sectional study of Chinese middle-aged and elderly populations; hence, the results cannot prove a causal relationship between PP and AoD in patients with HFrEF.The generalizability of our findings to the general population or other ethnic groups may be limited.Therefore, prospective studies with larger sample sizes are required to further explore causality.Second, we could not completely account for residual confounders because of unmeasured or unknown variables.For example, medications such as antihypertensive and antidiabetic drugs may affect AoD (50,51).Additionally, the detailed molecular mechanisms supporting our findings remain elusive; therefore, laboratory-based experiments are required to gain insight into the association between PP and AoD.Finally, due to our lack of measurements, there is no central pulse pressure data.This is a retrospective study, and it is true that the central hemodynamic study data is more revealing and convincing.In fact, the population in our study is mainly about heart failure patients, and the central hemodynamic test is an invasive examination, which is not beneficial to the patients with heart failure, and only a few research institutions can carry out it in clinical practice.The results we obtained can attract the attention of qualified researchers for further validation, so that they can be generalized with simple clinical indicators.Indeed, there is a physiological difference between central artery pressure and peripheral artery pressure, that is, the physiological amplification of PP (52).It is well known that the degree of pulse pressure amplification is strongly associated with total mortality and major cardiovascular events (26).

Conclusions
Our findings indicated that PP level was independently and positively associated with AoD, especially in older patients with HFrEF, but not in patients with HFmrEF and HFpEF.Arterial stiffening or vascular aging may play a certain role in the older HFrEF patients.

FIGURE 3
FIGURE 3There was no non-linear association between PP and AoD assessed by restricted cubic spline analysis in patients with HFrEF (p−non −linear = 0.538, p−overall < 0.001).

TABLE 2
Multivariate linear regression analysis for AoD and AoD/BSA.

TABLE 3
Comparison of clinical characteristics of patients with hFrEF in the quartile of PP levels.

TABLE 4
The association between PP levels and aortic dilation in HFrEF.Model 1 was adjusted for age, pulse pressure, gender, and body surface area.Model 2 was adjusted for covariates in model 1 plus smoking status, hypertension.Model 3 was adjusted for covariates in model 2 plus cardiac index (left atrial diameter and ejection fraction), hematological index, diabetes mellitus, coronary heart disease, chronic kidney disease, and stroke.
In a community-based cohort (n = 3,108), Daisuke et al. (22) observed that participants in the highest AoD quintile (35.2 ± 2.8 mm) had higher PP than those in the lowest quintile (30.8 ± 2.8 mm).Recently, a study in Korea demonstrated a significant positive association between invasively measured aPP and AoD/BSA