Diastolic dysfunction, blood pressure and obesity: new insights from a general population.
In this issue of the Journal, Millen et al.[1] report the results of a cross-sectional population-based study aimed at assessing the impact of blood pressure (BP) and clinical indexes of obesity on left-ventricular diastolic function, as assessed by conventional and tissue Doppler imaging (TDI) parameters, in a cohort of 417 black African individuals aged above 16 years (35% men), free of overt cardiovascular disease, but with high prevalence of overweight/obesity. Before addressing the results of the study, available evidence on this issue and related topics may warrant some considerations. Nowadays, congestive heart failure is the most common cause of hospitalization in developed countries, among major forms of cardiovascular disease, followed by cerebrovascular and coronary heart disease; hospitalization rates of congestive heart failure is also continuously increasing worldwide [2]. Congestive heart failure related to left ventricular diastolic dysfunction is an emerging clinical entity involving up to 60% of patients managed in clinical practice [3]. The likelihood of diastolic dysfunction as a dominant cause of heart failure is particularly high in patients with systemic hypertension, left ventricular hypertrophy (LVH), central obesity, type 2 diabetes mellitus, metabolic syndrome, sleep apnea syndrome and permanent atrial fibrillation [4]. Echocardiography is recommended by international guidelines and authoritative scientific bodies for diagnosing heart failure with preserved ejection fraction and identifying early diastolic abnormalities in asymptomatic patients at high cardiovascular risk [5,6]. A large amount of evidence indicates that the assessment of diastolic function by using conventional Doppler and TDI may provide relevant predictive information about cardiovascular morbidity and mortality independently from left ventricular mass, left ventricular geometry and Framingham score [7]. Among the conventional Doppler indexes assessing diastolic function, the ratio of the peak early (E) to peak late (A) mitral inflow velocity (E/A ratio), and deceleration time of the early mitral flow have been widely used in clinical research and current practice. In particular, the relation of E/A ratio with cardiovascular outcomes has a U-shaped form, the lowest and highest values being associated with a poor cardiovascular prognosis in different clinical settings. In the Cardiovascular Health Study [8], the first prospective survey showing that diastolic dysfunction is predictive of incident congestive heart failure, the E/A ratio was related to the relative risk of incident heart failure according to a U-shaped curve; in particular, for E/A values less than 0.7, the risk was 1.9 and for ratios greater than 1.5, the risk was up to 3.5 compared to intermediate E/A values. In the past decade, the TDI technique has been extensively validated in a variety of cardiac diseases. At present time, TDI, as well as new imaging technologies such as strain rate and speckle tracking, offer incremental information on myocardial function compared to conventional echocardiography. It has been shown that invasive measures of diastolic function better correlate with early diastolic TDI velocity than with early mitral flow velocity or E/A ratio. A reduced early myocardial velocity (Ei) velocity in patients with impaired left ventricular systolic function as well as in those with hypertension and LVH has been reported to predict mortality better than clinical and conventional echocardiographic data [9]. Furthermore, the E/Ei ratio reliably correlates with left ventricular filling pressure and is a strong independent predictor of fatal and nonfatal cardiovascular events [10]. The prevalence and correlates of diastolic dysfunction have been extensively investigated in a variety of clinical settings and in general population samples, because the identification of risk factors and mechanisms leading to subclinical left ventricular diastolic alterations is important to identify patients at risk of developing heart failure. Most attention has been focused on hypertension and obesity, due to the high prevalence of these modifiable risk factors in the general population. The Assessment of Prevalence Observational Study of Diastolic Dysfunction (APROS diadys) project, a cross-sectional observational study on elderly hypertensive patients (mean age 70 years, 49% men) without systolic dysfunction consecutively attending hospital outpatient clinics in Italy, evaluated the prevalence and clinical correlates of echocardiographic signs of diastolic dysfunction [11]. Diastolic dysfunction (E/A ratio <0.7 or >1.5) was found in 649 out of the 2425 patients (25.8%). Multiple logistic regression analysis found age, sex, left ventricular mass, systolic and pulse pressures, and mid-wall shortening fraction, but not BMI, as significant covariates. In the Flemish Study on Environment, Genes, and Health Outcomes (FLEMENGHO), 539 participants (49% men, mean age 52 years), selected from a random population sample stratified by sex and age, underwent a comprehensive echocardiographic and Doppler evaluation including TDI measurements [12]. The overall prevalence of left ventricular diastolic dysfunction was as high as 27.3% and increased in frequency with age. Mild diastolic dysfunction or impaired relaxation (abnormally low age-specific trans-mitral E/A ratio and normal left ventricular filling pressure as defined by an E/e′ ratio <9), moderate (elevated left ventricular end-diastolic filling pressure with E/A ratio within the normal age-specific range) and severe diastolic dysfunction (elevated left ventricular end-diastolic filling pressure with abnormally low E/A) were found in 9.8, 14.1 and 3.4%, respectively. Compared with individuals with normal diastolic function, those with moderate-to-severe diastolic impairment had a significantly higher age, BMI, heart rate and SBP. In that study, the association between obesity and diastolic dysfunction was clearly supported by multivariate regression full-adjusted analyses, showing that the risk of diastolic dysfunction increased by 2.5-fold for each 5 kg/m2 increment of BMI. In the Baltimore Longitudinal Study on Aging, a cross-sectional survey involving 843 healthy participants (mean age 67 years, 45% men), the authors investigated the relationship between different fat depots, assessed by dual-energy X-ray absorptiometry and abdominal computed tomography, and left ventricular diastolic function [13]. They found that increased visceral adiposity was independently associated with left ventricular diastolic dysfunction, whereas this was not the case for subcutaneous adiposity. These findings suggest that the redistribution in the pattern of adiposity, due to disproportionate increase in visceral fat as opposed to subcutaneous fat, exerts a detrimental effect on diastolic function. In the paper published by Millen et al.[1], the authors assessed the influence of clinic BP, waist circumference and BMI on indexes of diastolic function in a large cohort of middle-aged asymptomatic black African individuals (mean age 45 years). Only a marginal fraction (9%) of this population-based sample was normal-weight, 68% were obese, 47% were hypertensive, and 27% had type 2 diabetes or an glycated haemoglobin above 6.1%. As many as 70 individuals (16.6%) were found to have echocardiographic LVH (i.e. left ventricular mass index >51 g/m2.7). Diastolic dysfunction was identified in approximately one-third of the entire sample; of note, mild dysfunction (impaired relaxation with E/A ratio <0.75) was present in 13%, moderate dysfunction (E/A >0.75 <1.5 and E/e′ ratio >10) in 12%, and severe dysfunction in 4% (E/A >1.5 and E/e′ >10), respectively. The main results of this study can be summarized as follows: waist circumference, at variance from BMI and waist-to-hip ratio, was independently correlated to E/A and E/e′ ratio, considered as continuous variables; this relation, however, failed to maintain a statistical significance when these indexes of diastolic dysfunction were analysed as categorical variables (i.e. presence or absence of mild or moderate-to-severe diastolic dysfunction); SBP emerged as the most important correlate of diastolic dysfunction, assessed as a continuous or categorical variable; the strength of the association of SBP and diastolic dysfunction remained highly significant even after inclusion of left ventricular mass index in regression models. Some aspects and limitations of this interesting survey deserve to be briefly discussed. First, the lack of an independent association between obesity and diastolic dysfunction found in their series of middle-aged participants was ascribed by the authors to the possibility that the excess of adiposity may impact negatively on diastolic function only in conjunction with advanced age. This hypothesis, however, is challenged by a number of observations made in young overweight/obese individuals. For instance, Dhuper et al.[14], looking at the prevalence and clinical correlates of left ventricular geometric patterns and their relation to cardiac function in a sample of 213 obese African-American patients (mean age 14 years) compared to 130 normal-weight, age-matched counterparts, found that obesity, hypertension and concentric hypertrophy were independent predictors of diastolic dysfunction. The obese group had significantly higher left ventricular mass index, left atrium diameter and E/e′ ratio than the normal-weight group, whereas an opposite trend was noted for the E/A ratio. More in general, it has been shown that isolated obesity (i.e. without LVH, hypertension or type two diabetes mellitus) in young or middle-aged individuals is frequently associated with mild diastolic dysfunction [15]. Second, the results of the study by Millen et al., which show that neither waist circumference nor BMI were independently correlated to mild or moderate-to-severe diastolic dysfunction, cannot be regarded as an incontrovertible proof of a negligible impact of obesity on diastolic function in this general population. Indeed, waist circumference had an independent relationship with continuous variables of diastolic function such as E/A and E/e′ ratio. This finding points towards an association between central adiposity and left ventricular diastolic performance when diastolic indexes are expressed in a continuous way rather than using a dichotomous (and somewhat arbitrary) categorization. Third, recently developed echocardiographic techniques, such as three-dimensional echocardiography and speckle-tracking imaging, provide a more detailed insight into the left ventricular systolic and diastolic mechanics in all three spatial directions. It has been reported that the capability of these innovative techniques in detecting early left ventricular functional alterations is greater than that provided by conventional echocardiography and their use may allow to identify subtle mechanic left ventricular abnormalities even in patients with normal cardiac function as assessed by routine ultrasound examinations [16,17]. In obese patients, the likelihood of impaired left ventricular mechanics (longitudinal, circumferential and radial) is particularly high, due to a variety of adverse hemodynamic and nonhemodynamic factors impacting on myocardium [18,19]. Thus, data on prevalence and correlates of diastolic dysfunction in obesity based on traditional echocardiographic and Doppler parameters, including TDI, may substantially underestimate the real, adverse impact of this condition on the heart. Fourth, a novel information provided by Millen et al.[1] is that SBP clearly emerges as the most important factor associated with E/e' ratio (a validated index of left ventricular filling pressure) in a middle-aged general black African population with a high prevalence of obesity. As stated by the authors, a practical implication of this relevant observation is that an effective antihypertensive treatment may be sufficient to prevent the evolution from subclinical diastolic dysfunction to overt heart failure, independently from lifestyle measures aimed to achieve weight loss. The generalizability of this assumption is disputable. Prevalence rates of adequate BP control in hypertensive patients with persistent obesity are markedly lower than in their counterparts undergoing effective weight loss programs. Rapid weight loss induced by bariatric surgery exerts important cardio-protective effects in the morbid obese through LVH regression, improvement in left ventricular geometry and diastolic function, and reduction in left atrial size, irrespective of the pretreatment BP status and of the magnitude of the BP changes following bariatric procedures [20]. In summary, despite some intrinsic and perhaps unavoidable limitations, the report by Millen et al.[1] provides new data on the 'relative weight' of hemodynamic versus anthropometric variables in causing alterations in diastolic function observed in a general population. Further research (not so easy to be carried out, however) will be needed to confirm the results of the present study by investigating the relative impact of blood pressure and body weight-lowering intervention on the diastolic function. ACKNOWLEDGEMENTS Conflicts of interest There are no conflicts of interest.
Read more
12