Introduction

Abdominal obesity (AO) is an independent and modifiable cardiovascular risk factor related to a higher occurrence of coronary artery disease, left ventricular (LV) dysfunction and heart failure.1, 2, 3 Obesity and arterial hypertension (AH) are closely related and their coexistence results in the harmful impairment of cardiovascular structure and function.1, 4, 5 However, previous studies suggest that overall obesity may result in different hemodynamic alterations than AH per se.5, 6 Normotensive obese subjects were shown to have higher volume load, higher cardiac output and lower systemic vascular resistance.6, 7 However, non-obese hypertensives present with vasoconstriction as the main mechanism of increased blood pressure (BP).5 These cardiovascular alterations in obese hypertensives can overlap, resulting in various hemodynamic constellations.5, 6, 7, 8 This multifaceted interplay becomes even more complicated in view of the findings that patients with AO may present different hemodynamic patterns than those with overall obesity.8 AO correlates with higher metabolic activity strongly related to the increased activity of the sympathetic nervous system and the renin–angiotensin–aldosterone system, as well as altered endothelial dysfunction and hyperinsulinemia.8, 9, 10, 11 As a consequence, hypertensives with android obesity present higher arterial stiffness and impaired LV performance.11, 12, 13 It is suggested that AO should be included in the models assessing cardiovascular risk as an independent variable.14, 15

These hemodynamic differences between the types of obesity and AH provoke vivid scientific discussions. There is still a need for research on these complex mechanisms to gain a better understanding of the pathophysiological background of the obesity-related AH, which is essential for development of successful treatment strategies. Although the use of noninvasive techniques sensitive enough to detect subclinical cardiovascular disturbances is still limited, some promising methods appear to be on the horizon. Global longitudinal systolic strain (GLSS), a modern, well-validated, echocardiographic method of assessment of LV contractile deformation,16, 17, 18 was shown to be an independent predictor of outcome in patients with AH, heart failure and myocardial infarct.19, 20 Other noninvasive diagnostic techniques, such as impedance cardiography (ICG) and applanation tonometry (AT), provide additional data on arterial stiffness, central BP, fluid load and LV performance.21, 22 When used together, these methods provide detailed insight into the sex-specific patterns of aging23 and the pathophysiological background of subclinical LV dysfunction.24

Based on these assumptions, we hypothesize that at the early stage of hypertensive disease related to AO, the complex ventricular–vascular interplay may be successfully assessed by the combination of the above methods. Thus, the purpose of this study was to investigate the relationship between AO and hemodynamics that was evaluated with the use of echocardiography, AT and ICG in young and middle-aged hypertensives.

Methods

Study population

The group selected for this analysis comprised patients with at least a 3-month history of AH defined according to the European Society of Cardiology guidelines1 and who were included in the prospective clinical study. We excluded patients with the following criteria: (1) confirmed secondary AH, (2) AH treated with three or more medicines before recruitment, (3) heart failure, (4) cardiomyopathy, (5) significant heart rhythm disorders, (6) significant valvular disease, (7) kidney failure (glomerular filtration rate below 60 ml*min−1 per 1.73 m2), (8) chronic obstructive pulmonary disease, (9) diabetes, (10) polyneuropathy, (11) peripheral vascular disease or (12) an age <18 years. The subjects treated with hypotensive drugs were recommended to discontinue them at least 7 days before the examination. The study was conducted according to the Good Clinical Practice guidelines and the Declaration of Helsinki, with the approval of the local ethics committee. Each patient provided written informed consent to participate in the study.

The clinical examination included analysis of cardiovascular risk factors, symptoms indicating secondary cause of AH and laboratory tests (creatinine, glomerular filtration rate, total cholesterol, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, triglycerides and fasting glucose). Body mass index was derived from height and weight. Patients with metabolic syndrome were defined according to the IDF (International Diabetes Federation) criteria25 and AO according to the European Society of Cardiology guidelines as waist circumference (WC) in men ⩾102 cm and in women ⩾88 cm.1

Office blood pressure measurement

The office blood pressure measurement (OBPM) was performed in the morning (Omron M4 Plus, Omron Healthcare Co. Ltd., Kyoto, Japan) by a technique compliant with the European Society of Cardiology guidelines.1 Automated systolic BP (SBP) and diastolic BP (DBP) measurement was supervised by a trained physician or nurse, after a minimum 5 min of rest in a sitting position. The study only included the patients with AH confirmed in ABPM.

Echocardiography

Two-dimensional echocardiography was performed using standard parasternal, apical and subcostal views (2.5 MHz transducer; VIVID S6 GE Medical System, Wauwatosa, WI, USA). The dimension of the left atrium (LA), left ventricular end-diastolic diameter (LVEDD) and interventricular septum diameter were measured in the parasternal long-axis view in late diastole of the LA and left ventricle (LV), respectively. Left ventricular enlargement was defined as LVEDD indexed to the body surface area >3.2 cm m−2 for men and >3.1 cm*m−2 for women, and LA enlargement as LA >4.0 cm for men and >3.8 cm for women. Left ventricular ejection fraction was calculated according to Simpson’s formula using a two-dimensional image of the LV chamber during systole and diastole in the 4- and 2-chamber apical views. The left ventricular hypertrophy (LVH) was diagnosed according to the ASE-recommended formula for estimation of the left ventricular mass index (LVMI) from two-dimensional linear LV measurements and indexed to the body surface area (cutoff values for men LVMI >115 g*m−2 and for women >95 g*m−2).

Mitral valve inflow was recorded in the apical 4-chamber view with pulsed wave Doppler gate positioned in the LV on the level of the mitral valve edges. The following parameters were measured: mitral inflow early (E) and late (A) phase ratio (E/A), and early phase deceleration time. The apical 5-chamber view enabled simultaneous registration of the inflow pattern through aortic and mitral valves and isovolumic diastolic time calculation. Tissue Doppler imaging was performed in the apical views to acquire mitral annular velocity. The sample volume was positioned within 1 cm of the septal insertion sites of the mitral leaflets and adjusted as necessary (usually 5–10 mm) to cover the longitudinal excursion of the mitral annulus in diastole. Additionally, mitral septal annulus early diastolic velocity (e′) was measured and the E/e′ ratio was calculated. The diagnosis of left ventricular diastolic dysfunction (LVDD) was based on the current guidelines.26 The study group only included patients with normal LVD function and LVD dysfunction stage I (impaired relaxation diastolic filling pattern, which is related to the reduced left ventricular filling in early diastole). The following values were considered abnormal: e′ <8 cm*s−1; LA >40 mm for men and >38 mm for women; E/A <0.8; early phase deceleration time >200 ms; isovolumic relaxation time ⩾100 ms; and E/e′ ratio >8.

For the assessment of GLSS, the dedicated automated function imaging protocol was used. Digital images acquired in the apical long-axis, apical 2-chamber and apical 4-chamber and loop recorded with ECG gating were analyzed. As required, high temporal resolution of >50 frames per second was obtained to enable acoustic myocardial marker tracing. The detection of the tracked area was carried out semiautomatically after selection of two basal points at the level of mitral annulus and the third point in the apex, with manual correction when needed. In each of the apical views, LV walls were divided into six segments. The value of strain and quality of tracing were then assessed for each LV segment. The mean longitudinal peak systolic strain value was calculated for each of the three views. The value of the GLSS was calculated as the arithmetical mean of these values.

Impedance cardiography

All ICG measurements were performed using a Niccomo device (Medis, Ilmenau, Germany) after 10 min of rest in a supine position. The data were recorded during a 10-min assessment and exported to the dedicated software (Niccomo Software, Medis). The final analysis included mean values of hemodynamic parameters characterizing (1) cardiac performance: stroke volume (SV) and its index to body surface area (SI), cardiac output (CO) and its index (CI), acceleration index (ACI), velocity index (VI) and Heather index (HI); (2) afterload: systemic vascular resistance (SVR) and its index, and total arterial compliance (TAC); and (3) preload: thoracic fluid content (TFC) and its index (TFCI). The calculation formulas were presented in our previous publication.27

Applanation tonometry

The assessments of the central BP (CBP) and augmentation index (AI) were performed noninvasively using the SphygmoCor system (AtCor Medical, Sydney, NSW, Australia). Radial artery pressure waveforms were recorded at the left wrist using AT with a high-fidelity micromanometer (Millar Instruments, Houston, TX, USA). The arterial pulse waves were processed with SphygmoCor software (version 9.0; AtCor Medical Inc. Pty Ltd, Sydney, Australia), and the corresponding aortic pressure waveform was generated from the radial artery waveform using a validated transfer function with the identification of an inflection point resulting from the wave reflection and the incisura resulting from the aortic valve closure.28 As a result, central systolic BP (CSBP), central diastolic BP (CDBP) and central pulse pressure (CPP) were derived. Augmentation pressure (AP) was calculated by the maximum systolic pressure minus the pressure at the inflection point and augmentation index (AI as AI=AP × 100/CPP). Only high-quality recordings (quality index >80%) were included in the analysis. The measurements were performed in the supine position just after examination by ICG. The radial pulse and transferred aortic blood pulse were calibrated against the last measurement of brachial SBP and DBP by the oscillometric module of the Niccomo device.

Statistical analysis

The statistical analysis was performed using Statistica 7.0 software (StatSoft Inc.,Tulsa, OK, USA). The distribution and normality of the data were assessed by visual inspection and the Kolmogoro–Smirnov test. Continuous variables were presented as the means±s.d., and categorical variables were presented as absolute and relative frequencies (percentages). To analyze the differences between subgroups of patients with and without AO, Student’s t-test was used for data with normal distribution, and the Mann–Whitney U-test was used if the data were not normally distributed. For categorical variables, the χ2 test and the Fisher's exact test were used. The associations between clinical features, indices of the cardiovascular function and structure with WC were analyzed with Pearson’s correlation coefficients. A P-value <0.05 indicates statistical significance.

Results

Basic characteristics

We identified AO in more than half of the patients (n=79, 54.9%, Table 1). AO was associated with slightly higher office BP, but age, HR, GFR and gender distribution were similar between patients with and without AO (Table 2).

Table 1 Basic characteristics of the study group
Full size table
Table 2 The comparison of the clinical and echocardiographic characteristics in patients with and without abdominal obesity
Full size table

Echocardiographic assessment

AO correlated with larger left chambers and aortic dimensions (LVEDD, LA, AoA), lower absolute value of GLSS and higher prevalence of LVDD (Table 2). No significant differences for left ventricular ejection fraction, LVMI and the prevalence of LVH were observed.

ICG and AT assessment

The most prominent differences in ICG assessment were observed for indices of LV performance and thoracic fluid load. Obese subjects presented with lower values of LV contractility (VI, ACI, HI) but normal SI and CI (Table 3). Moreover, obesity was associated with lower TFC, especially if indexed to the body surface area. No significant differences for afterload parameters (TAC, systemic vascular resistance index, CPP and AI) were observed. However, the consistent trend toward higher central arterial stiffness (lower TAC, higher CPP and AI) seems to be clinically relevant.

Table 3 The comparison of hemodynamic characteristics in patients with and without abdominal obesity
Full size table

Correlation analysis

In the analysis of the association between AO and other analyzed variables, the most prominent correlations were noted for WC, LVEDD, LA, AoA, GLSS, E/A, VI, ACI, HI and TFCI (Tables 2 and 3; Figure 1). Statistically borderline relationships, not revealed in intergroup comparison, were noted for WC with SBP, DBP, LVMI and RVEDD.

Figure 1
figure 1

The correlation plots of waist circumference with: left heart dimensions (a, left ventricular end-diastolic diameter (LVEDD); b, left atrium diameter (LA)); indices of the left ventricular performance (c, global longitudinal systolic strain (GLSS); e, velocity index (VI); f, Heather index (HI)) and mitral inflow pattern (d, mitral inflow early and late phase ratio (E/A)).

Full size image

Discussion

The results of our study confirmed the relationship between AO and cardiovascular function. The use of complementary diagnostic methods revealed that the excessive accumulation of abdominal fat even in young and middle-aged hypertensives is related to complex hemodynamic alterations. Our results confirmed previous findings in this area and revealed new insights into the pathogenesis of obese-related AH. We showed that asymptomatic impairment of cardiac performance could be the earliest detectable clinical feature of the impaired cardiovascular function related to AO. Both LV diastolic and systolic disturbances were more prominent than alterations in volume load and afterload. Our results substantiate aggressive diagnostic and therapeutic strategies in obese hypertensive patients and indicate the use of easy diagnostic tools, which are promising in preliminary evaluation and the monitoring of the intervention.

AO and left ventricular diastolic function

Our results confirmed a significant association between AO and the impaired LV diastolic function. We observed that the presence of AO favored LVDD, which was exhibited in the impaired LA–LV interaction (lower E/A), prolonged LV relaxation (longer isovolumic relaxation time) and increased LV filling pressure (lower e′ and higher E/e′). The accelerated impairment of LVDD in obese hypertensives was reported by Miyoshi et al.29 These findings also agree with several previous observations30, 31 reporting that an expanded blood volume, LV dilation30 and LVH30, 31 may contribute to LVDD. However, in our study, the prevalence of LVH was very low and there was no case of LV enlargement. In this context, the other factors should be considered as responsible for early-stage LVDD. Parrinello et al.32 illustrate the crucial role of metabolic and neurohormonal abnormalities that are especially prominent in AO. Left ventricular stiffening and hypertrophy are probably two parallel processes induced by one or more of the following: activated sympathetic nervous system, renin–angiotensin–aldosterone system, oxidative stress, insulin resistance, hyperleptinemia, myocardial fibrosis and intramyocellular triglyceride deposition.11, 31

AO and left ventricular systolic function

The mean GLSS (−17.2%) observed in obese hypertensives was higher (lower absolute value) compared with that in the normotensives (−21.1%,17 −20.2%18). The harmful effect of obesity on LV contractility was previously reported. Wong et al.30 found significant differences in myocardial contractility between obese and lean patients. They also showed that the degree of LV dysfunction correlates with body mass index. Obesity in diabetic patients was related to the impaired LV systolic function and deformation in three-dimensional speckle tracking echocardiography. Moreover, the coexistence of overall obesity and AO resulted in the most impaired myocardial systolic performance.33 Obert et al.34 examined asymptomatic, severely obese adolescents and observed significantly altered longitudinal systolic and early diastolic strain rate in the presence of preserved circumferential deformation. This particular sensitivity of GLSS is associated with specificity of myocardial fibers, which are responsible for longitudinal deformation. The fibers are placed subendocardially and therefore are more exposed to ischemia, fibrosis and impaired microcirculation.11

The coexistence of lower GLSS with a higher prevalence of LVDD in obese patients suggests a strong association of these disturbances. These interrelations were previously reported in patients with AH.35, 36 Galderisi et al.17 observed that GLSS was an independent contributor to E/e′ ratio (P<0.0001) separate from age, heart rate, meridional end-systolic stress, LV mass index and left atrial volume index. Additionally, Ballo et al.37 proved that GLSS was superior to circumferential strain in predicting LVDD and increased LV diastolic pressure (E/e′ ⩾13).

The novelty of our approach lies in the hemodynamic assessment of LV performance by ICG and AT. Similar to previous reports,8, 38 we observed higher SV and CO in obese subjects, but after normalization to body dimensions, the values of SI and CI were comparable between obese and non-obese subjects. The significantly depressed parameters, which characterize LV performance as a blood pump (ACI, VI, HI), were found to be useful markers of the obesity-related hemodynamic alterations. The impaired dynamics of LV outflow may reflect reduced LV contractility as with impaired capacity of the aorta to counterbalance forward (blood ejection) and backward (afterload) load forces. These findings provide new insight into the pathogenesis of the obese-related exercise intolerance. Impaired LV contractility and blood outflow may result in the insufficiency of circulatory supply in the case of increased body demand during exercise.

AO and left ventricular load

The differences between hemodynamic indices of afterload have not reached statistical significance in our analysis, but the consistency of the evaluation by two independent methods (ICG and AT) suggests clinically credible proof of increased arterial stiffness in the subgroup with AO. Lower TAC, higher CPP and AI indicate functional stiffening of the aorta, with limited impact of peripheral resistance (systemic vascular resistance index).39 We confirmed that hemodynamic alterations overlapped in obese hypertensives. The possibility to assess afterload may be clinically important in view of the discrepancies between studies reporting arterial status in obesity and AH.3, 5, 6, 13, 40 These apparently confusing observations reflect the interplay between hypertension and obesity. Healthy, normotensive, obese patients present with low peripheral vascular resistance,3, 6, 40 but the superposition of even borderline AH moderates this potentially positive effect.5, 6 Kangas et al.13 observed that arterial stiffness (pulse wave velocity, aortic pulse pressure and AP) is more pronounced in patients with metabolic syndrome, particularly with hypertension. In a prospective observation, Safar et al.4 showed that metabolic disorders increase the risk of the accelerated vascular aging. Rizzoni et al.7 explain that both obesity and AH influence artery structure resistance but with different mechanisms; namely, metabolic disturbances promote vascular smooth muscle growth (hypertrophic remodeling), whereas AH has an increased media-to-lumen ratio (eutrophic remodeling). The obesity-related endothelial dysfunction is another potential mechanism of the impaired vasodilation.9 This imbalance between high output and vasoconstriction may contribute significantly to the increased BP. The observations of Fujii et al.41 emphasize the role of inflammation in hemodynamic disturbances related to AO. The authors followed 705 normotensive subjects and reported increased risk of new-onset hypertension related to copresence of AO and a high level of high-sensitivity C-reactive protein. Kim et al.42 demonstrated that AO is associated with early metabolic disturbances and renal impairment showing a significant correlation between visceral adipose tissue (evaluated by computed tomography), insulin resistance and prevalence of microalbuminuria.

Relatively low TFC/TFCI, as we have observed in the AO subgroup, is quite difficult to interpret. One might expect an increased TFC/TFCI as a reflection of obesity-related fluid retention. However, in our opinion, TFC should be regarded as an indicator of the local fluid content but not the overall body fluid balance. Low TFC may suggest that young and middle-aged obese hypertensives, even with slightly impaired LV performance, do not present pulmonary congestion. Presumably, the LV is efficient enough to cope with the volume load at this preliminary stage of dysfunction. Another issue is that AO may contribute to venous stasis in lower body compartments by mechanical compression of the abdominal veins,43 which may partly explain the relatively low accumulation of fluid in the thorax (TFC/TFCI). However, in our opinion, previous results have not provided enough evidence to consider TFC/TFCI as a good marker of preload in AO. Whole-body impedance methods could provide additional value in the assessment of fluid balance and distribution in obese patients.

AO and heart structure

We observed that obese patients were associated with larger dimensions of the left heart chambers. These anatomic differences were previously reported2, 5, 12, 29, 38 and will not be further discussed. We would like simply to comment that higher LVDD, LA and AoA should be interpreted not only as a reflection of overall body dimensions but also of altered LA–LV interaction and central hemodynamics. Left ventricular mass index was slightly higher in hypertensives with AO and correlated with WC. The difference in the prevalence of LVH was not statistically significant but agreed with clinical trends of higher occurrence of myocardial hypertrophy in obese subjects.2, 5, 6, 38

Clinical implications

The combination of AO and AH is a double burden to the heart. We identified LV dysfunction as the prominent marker of the impaired hemodynamics in obese hypertensives. We show that echocardiographic assessment may be supported by noninvasive diagnostic tools (ICG and AT) and that it may be easily used in everyday practice. These methods revealed the complex interplay between obesity-related (volume load) and hypertension-related (arterial stiffness) hemodynamic alterations, which influence LV performance. The identification of individual hemodynamic profiles may support personalized pharmacotherapy and provide an objective evaluation of the response to the therapeutic intervention. Such an approach may result in better treatment outcomes, including reverse myocardial remodeling and preserved cardiac function.

Limitations

We are aware that the small sample size is a limitation of the study. The use of noninvasive techniques is also biased by non-direct hemodynamic measurements that may be altered by the patient’s temporary state (BP, heart rate, preload and afterload). However, alternative techniques are either invasive or more expensive and cannot be widely used. Another limitation is that we did not perform analyses of echocardiographic radial and circumferential deformation. We also did not exclude asymptomatic ischemic heart artery disease (i.e., by performing an exercise test and/or coronarography). However, no signs or symptoms suggesting cardiac ischemia, such as LV wall motion abnormalities, electrocardiogram findings, chest pain and/or dyspnea, were noted. Previous hypotensive treatment should also be considered as a potential limitation, but it concerned <20% of subjects. Moreover, our study comprised mostly young and middle-aged hypertensives and our conclusions should not be extrapolated to the general population. However, the strength of these results is the population specificity for AH and lack of potential bias related to additional cardiovascular alterations.

Conclusions

The impact of AH and AO on cardiovascular hemodynamics overlaps and, at the early stage, is exhibited in impaired cardiac performance. The asymptomatic depression of the left ventricular diastolic and systolic function seems to be the earliest clinical feature of impaired ventricular–vascular interactions. The assessment of the individual hemodynamic profile with the use of modern noninvasive diagnostic methods should be considered in personalized therapy that aims at the prevention of adverse cardiovascular events.