Opposite Associations of Circulating Aldosterone and Atrial Natriuretic Peptide with Left Ventricular Diastolic Function in Essential Hypertension

Home , Plasma renin activity

ORIGINAL ARTICLE Opposite associations of circulating aldosterone and atrial natriuretic peptide with left ventricular diastolic function in essential hypertension

RH Fagard, PJ Lijnen and VV Petrov Hypertension and Cardiovascular Rehabilitation Unit, Department of Molecular and Cardiovascular Research, Faculty of Medicine, University of Leuven KU Leuven, Leuven, Belgium

It has been shown in animal experiments that angioten- (P Ͻ 0.001) and to plasma aldosterone (P Ͻ 0.01), and sin II and aldosterone have mitogenic effects on the car- positively to plasma atrial natriuretic peptide (P Ͻ 0.05). diovascular system, whereas atrial natriuretic peptide Peak flow velocity during atrial contraction was posi- has antimitogenic properties. The aim of the present tively related to plasma atrial natriuretic peptide both study was to relate plasma renin activity, angiotensin II, before (P Ͻ 0.001) and after (P Ͻ 0.05) controlling for aldosterone and atrial natriuretic peptide to left ventricu- significant covariates (age, sex and blood pressure). We lar structure and function, assessed by use of imaging conclude that circulating renin, angiotensin II, aldos- echocardiography and transmitral Doppler velocimetry terone and atrial natriuretic peptide are not indepen- in 73 patients with essential hypertension, World Health dently related to left ventricular mass in essential hyper- Organization stages I–II, aged 43 ؎ 10 (s.d.) years. Left tension. The inverse association of plasma aldosterone ventricular mass, wall thickness and internal diameter with indices of diastolic function is compatible with a were not independently related to the biochemical vari- stimulating effect of aldosterone on myocardial fibrosis, ables, except for a weak and positive association of wall which is opposed by atrial natriuretic peptide. The -However, apparently conflicting positive association of this pep .(0.06 ؍ thickness with plasma aldosterone (P left ventricular early inflow peak velocity and deceler- tide with atrial peak velocity is most likely due to stimu- ation were independently and inversely related to age lation of its secretion by atrial involvement.

Keywords: aldosterone; angiotensin; atrial natriuretic peptide; echocardiography; left ventricular function; left ventricular hypertrophy; renin

Introduction because long-term administration of components of the renin-angiotensin-aldosterone-system and of atr- There is evidence from animal experiments that ial natriuretic peptide cannot be performed on ethi- angiotensin II and aldosterone exert mitogenic cal grounds. Therefore researchers have looked for effects on the cardiovascular system, whereas atrial 1–5 associations between cardiac structure and function, natriuretic peptide has antimitogenic properties. and plasma renin activity,13–21 angiotensin The development of cardiac fibrosis on long-term in II,15,17,20,22 aldosterone15,17–20,22,23 and atrial natriur- vivo administration of aldosterone in rats has gener- etic peptide.22,24–27 Another approach has been to ated considerable interest, particularly because of its study the cardiac effects of chronic treatment with seemingly independence of haemodynamic fac- 6–9 drugs which interfere with components of the renin- tors; however, in vitro studies are not in agree- angiotensin-system.28,29 Growth stimulation of myo- ment on the direct effect of aldosterone on collagen 10,11 cytes would lead to an increase in left ventricular synthesis by cultured cardiac fibroblasts. Atrial mass, whereas myocardial fibrosis would mainly natriuretic peptide on the other hand has been result in functional abnormalities, particularly dias- shown to inhibit the proliferation of cultured rat 3 12 tolic dysfunction. Imaging echocardiography is an fibroblasts, but it is not known whether long-term appropriate non-invasive technique to estimate left infusion of atrial natriuretic peptide would counter- ventricular mass, whereas mitral inflow Doppler act or reverse the fibrotic process. Such issues are of velocimetry allows assessment of diastolic func- paramount importance for human hypertensive tion.30 Moreover, echocardiographic indices of dias- heart disease, but are difficult to resolve in man, tolic function correlate with myocardial fibrosis in hypertensive man.31 Few studies have addressed the Correspondence: Dr R Fagard, UZ Pellenberg, Weligerveld 1, B- relationships of plasma atrial natriuretic peptide 3212 Pellenberg, Belgium with left ventricular characteristics,22,24–27 whereas Received 13 June 1997; revised and accepted 26 November 1997 the data on the renin-angiotensin-system, both from Hormones and ventricular diastolic function RH Fagard et al 196 the correlation approach and from intervention sional image, and were recorded on photographic studies are conflicting.13–23,28,29 In addition, most paper at paper speed 100 mm/s for later analysis. reports emphasized cardiac structure, whereas The transducer was angulated to ensure simul- relationships of humoral variables with diastolic taneous visualisation of the left ventricular posterior function have only been studied for atrial natriuretic wall and interventricular septum just below the peptide and not for the components of the renin- mitral valve apparatus, ie, at the level of the pos- angiotensin-system. The purpose of the present terior chordae tendinae. After completion of the study was therefore to assess relationships of echo- study the tracings were digitized and the values of cardiographic left ventricular structural and func- three to five heart beats were averaged. The rec- tional characteristics with plasma renin activity, ommendations of the American Society of Echocar- angiotensin II, aldosterone and atrial natriuretic diography were observed; wall and cavity measure- peptide in patients with essential hypertension, ments were made using the leading edge method.34 studied in well-standardized conditions. The following echocardiographic measurements were obtained: (1) left ventricular end-diastolic internal diameter (LVID; mm), measured at the onset Materials and methods of the QRS complex of the simultaneously recorded Protocol electrocardiogram, and end-systolic internal diameter at the nadir of septal motion (LVIDs; mm); and Previously treated or untreated patients, referred for (2) end-diastolic posterior wall and interventricular the evaluation of hypertension, were eligible for the septal thicknesses (PWT; IVST; mm). Several study when no underlying disease could be indexes were derived from these measurements: detected, when organ damage was limited to World (1) mean end-diastolic wall thickness (MWT; mm) Health Organization stages I or II32 and when their as the average of posterior wall and septal thick- echocardiogram was judged to be of good quality by nesses; (2) left ventricular mass (LVM; g) as two investigators. Patients willing to participate described by Devereux et al;35 (3) relative wall thick- were invited for two study visits within the next 6 ness (RWT) as the ratio of MWT to the internal weeks, with 2 to 3 weeks between visits; visits were radius (LVID/2); and (4) the fractional shortening of scheduled in the morning, after the patients had a the LV internal diameter (FS; %) as 100 × (LVID − light breakfast at home. Anti-hypertensive treatment LVIDs)/LVID. was interrupted in patients with mild-to-moderate hypertension. Patients with severe blood pressure elevation were only accepted when they were Pulsed Doppler echocardiography untreated and the two study visits were scheduled With the patient rotated on the left side, the trans- in a shorter period of time. No special diet instruc- ducer was placed at the apex and the apical four- tions were given and sodium intake was assessed chamber image was visualised. The Doppler sample from 24-h urinary sodium excretion. The study was volume was positioned in the orifice of the mitral approved by the Ethics Committee of the Faculty of valve and the ultrasound beam was placed parallel Medicine and participants gave informed consent. to the presumed flow of blood. The transducer was angulated to obtain the best Doppler signal, as judged by the sharpness of the outline of the com- Clinic blood pressure plexes on the visual display and the pitch of the After arrival in the laboratory the patients were left audio sound of the Doppler shift frequencies. Rec- alone in a quiet room, where they rested in the ordings were made at end-expiration. The peak velo- supine position. Twenty minutes later, blood pres- city (m/s) and deceleration (m/s2) of the early filling sure was measured five times by sphygmomanome- curve and the peak velocity of the atrial contraction try and auscultation (Korotkoff phases I and V) by curve (m/s) were measured on three to five heart one of two trained observers; pulse rate was assessed beats, using the middle of the darkest portion of the at the end of the blood pressure measurement by Doppler flow curve. The ratio of the early to atrial palpation of the radial artery during 30 s. The five peak velocity was calculated.36 blood pressures of each visit were averaged and mean blood pressure was calculated as: Biochemical measurements + diastolic pressure After termination of echocardiography, the patients − (systolic pressure diastolic pressure)/3. remained in the supine position for withdrawal of blood from an antecubital vein, for determination of 37 38 Imaging echocardiography plasma renin activity, angiotensin II, aldosterone,39 atrial natriuretic peptide40 and plasma nor- Echocardiography was performed by one of two adrenaline.41 operators under supervision of the same third inves- tigator, by use of an Ultra-Imager (Honeywell Inc, Statistical analysis Denver, CO, USA) or CV 750 device (Vingmed, Horten, Norway), with the subject rotated slightly on Database management and statistical analyses were the left side.33 M-mode echocardiograms of the left performed using SAS-software (SAS Institute Inc, ventricle were obtained at end-expiration from the Cary, North Carolina). Results of two visits were long axis view, under control of the two-dimen- averaged, except in two patients who were seen only Hormones and ventricular diastolic function RH Fagard et al 197 once. Group data are reported as mean (s.d.) or as Table 1 Echocardiographic data median and range. Positively skewed data were log- arithmically transformed, provided that transform- Imaging echocardiography Heart rate (beats/min) 65.3 ± 8.3 ation improved the skewness index. Results were Left ventricular mass (g) 231 ± 69 analysed using single and stepwise multiple Left ventricular mass index (g/m2) 123 ± 32 regression analysis. A two-tailed P value of р0.05 Left ventricular internal diameter (mm) 46.1 ± 4.9 was considered significant. Interventricular septal thickness (mm) 13.4 ± 2.8 Posterior wall thickness (mm) 12.4 ± 2.2 Mean wall thickness (mm) 12.9 ± 2.3 Results Relative wall thickness 0.57 ± 0.12 Fractional shortening (%) 31.3 ± 6.5 Characteristics of patients Transmitral Doppler echocardiography Age of the 73 patients (48 men, 25 women) averaged Early filling: Peak velocity (m/s) 0.56 ± 0.12 43.5 ± 10.1 years, weight 75.3 ± 12.6 kg, height 1.71 deceleration (m/s2) 4.13 ± 1.43 ± 0.09 m and body mass index 25.6 ± 3.2 kg/m2. Atrial filling: peak velocity (m/s) 0.46 ± 0.10 ± Blood pressure was 158 ± 19/102 ± 13 mm Hg and Ratio of early to atrial peak velocity 1.25 0.34 ± pulse rate 70.5 10.4 beats/min. Age, body mass ± index and blood pressure were not different Values are mean s.d. between men and women; women were smaller, weighed less than men and their pulse rate was hypertrophy and evidence of impaired left ventricu- higher (P Ͻ 0.05). Median duration of hypertension lar diastolic function.42 was 50 months (range 1.5–300). Forty-one patients had never been treated for hypertension or had not Determinants of biochemical variables received anti-hypertensive drugs for more than 1 month in the previous year. Treatment had been Plasma renin activity and angiotensin II were not interrupted for a median of 9 weeks and for at least significantly related to age, sex, height, weight, 3 weeks in all ever treated patients. Urinary sodium blood pressure, 24-h urinary sodium excretion, dur- excretion averaged 99 ± 51 mmol/24 h. Figure 1 ation of hypertension or use of previous medication. gives the median and the individual data for plasma Plasma aldosterone was positively related to mean renin activity, angiotensin II, aldosterone and atrial blood pressure (r = 0.24; P Ͻ 0.05) and plasma nor- natriuretic peptide, which were available in 72, 71, adrenaline to age (r = 0.23; P = 0.05). Single 71 and 72 patients, respectively, together with the regression analysis revealed that atrial natriuretic normal ranges in the authors’ laboratory. Plasma peptide was related to age (r = 0.47; P Ͻ 0.001), noradrenaline ranged from 0.66 to 7.9 ␮mol/L female vs male sex (r = 0.32; P Ͻ 0.01), mean blood (median: 2.04; n = 70). Plasma angiotensin II, atrial pressure (r = 0.33; P Ͻ 0.01), weight (r =−0.26; P natriuretic peptide and noradrenaline were logarith- Ͻ 0.05) and height (r =−0.36; P Ͻ 0.01); multiple mically transformed in all further analyses. Table 1 regression analysis showed independent contri- summarizes the results from imaging and Doppler butions of age (partial regression coefficient: +0.0105 echocardiography; when compared to normal log pmol/L/yr; P Ͻ 0.001), female sex (+0.146 log values, patients had concentric left ventricular pmol/L; P Ͻ 0.05) and mean blood pressure (+0.0045

Figure 1 Median (horizontal bar) and individual data for plasma renin activity, plasma angiotensin II, aldosterone and atrial natriuretic peptide concentrations. The vertical bars represent the normal ranges of the biochemical variables. Hormones and ventricular diastolic function RH Fagard et al 198 log pmol/L/mm Hg; P Ͻ 0.05) to atrial natriuretic terone (P Ͻ 0.001 (Figure 2), but to none of the other peptide. biochemical variables. The relationships between the early inflow characteristics and aldosterone remained significant (P Ͻ 0.001) after adjustment for Determinants of left ventricular structural significant covariates, that is age and sex for peak characteristics velocity (partial r =−0.39) and age for deceleration =− When age, sex, height, weight, mean blood pressure, (partial r 0.40). Whereas these inflow character- heart rate, 24-h urinary sodium excretion, duration istics were not significantly related to atrial natriur- of hypertension and previous medication were etic peptide in single regression analysis, these offered for selection in stepwise multiple regression relationships became significant after controlling for age (Figure 3); the partial correlation coefficients analysis, four significant variables explained 64% of Ͻ = the variance of left ventricula mass (P Ͻ 0.001), that amounted to 0.33 (P 0.01) and 0.29 (P 0.01), is mean blood pressure (partial regression coef- respectively. Table 3 illustrates the regression equa- ficient: +2.1 g/mm Hg; P Ͻ 0.001), weight (+2.7 g/kg; tions when all biochemical variables were offered P Ͻ 0.001), height (−210 g/m; P Ͻ 0.05) and sex for selection, together with demographic, anthropo- (−63 g for women vs men). The same variables metric, clinical and left ventricular structural data. explained 54% of the variance of mean wall thick- The results reveal opposite and independent associ- ness (P Ͻ 0.001). The left ventricular end-diastolic ations of plasma aldosterone and atrial natriuretic internal diameter was independently determined by peptide with the early transmitral inflow character- weight (+0.16 mm/kg; P Ͻ 0.001) and heart rate istics; these relationships remained significant when (−0.13 mm/bpm; P Ͻ 0.05) (R2 = 0.26; P Ͻ 0.001). sex was forced into the equations. There were no significant relationships of left ven- Transmitral peak velocity during atrial contrac- tricular mass, wall thickness and internal diameter tion was positively related to atrial natriuretic pep- with biochemical variables in single regression tide, both in single (Table 2) and in multiple analysis, except for mean wall thickness and plasma regression analysis (Table 3). The early to late peak aldosterone (r = 0.27; P Ͻ 0.05) (Table 2). None of flow velocity ratio was influenced by aldosterone the biochemical variables correlated significantly and atrial natriuretic peptide in single regression with the left ventricular structural characteristics in multiple regression analysis, when offered for selection together with the demographic, anthropometric and clinical characteristics. The positive relationship between mean wall thickness and aldosterone fell short of significance (P = 0.06) after controlling for mean blood pressure, weight, height and sex.

Determinants of left ventricular inflow characteristics The single correlation coefficients between left ventricular inflow characteristics and the biochemical measurements are given in Table 2. Peak velocity and deceleration of the early filling wave were significantly and inversely related to plasma aldos-

Table 2 Relationships between echocardiographic measurements and biochemical variables: single correlation coefﬁcients

LVM MWT LVID Epv Edec Apv EA ratio

PRA 0.03 0.13 −0.15 −0.15 −0.19 −0.01 −0.09 lANGII 0.07 0.14 −0.11 0.06 −0.04 0.11 −0.06 ALDO 0.15 0.27* −0.14 −0.42+ −0.44+ 0.06 −0.38+ lANP 0.04 0.05 −0.01 0.09 0.03 0.46+ −0.28* lNAPI 0.02 0.03 −0.03 0.05 −0.02 0.25* −0.15

*P р0.05; +P р0.001. LVM: left ventricular mass; MWT: mean wall thickness; LVID:

left ventricular internal diameter; Epv: peak velocity of early filling wave; Edec: deceleration of early filling wave; Apv: peak velocity of atrial filling wave; EA ratio: ratio of early to atrial peak velocity; PRA: plasma renin activity; lANG II: log plasma angio- Figure 2 Regression plots of peak flow velocity (upper panel) and tensin II concentration; ALDO: plasma aldosterone concentration; flow deceleration (lower panel) of the left ventricular early filling lANP: log plasma atrial natriuretic peptide concentration; lNAPI: wave (E) against plasma aldosterone concentration. The relation- log plasma noradrenaline concentration. ships remain significant after adjustment for age. Hormones and ventricular diastolic function RH Fagard et al 199 graphic indices of left ventricular diastolic function than to left ventricular structure in patients with essential hypertension; and (2) that a higher circulating aldosterone is associated with reduced diastolic function, whereas atrial natriuretic peptide shows the opposite relationship. Although angiotensin II and aldosterone have been shown to have growth stimulating and mitogenic effects on the heart in animal experiments, our cross-sectional findings could suggest that these properties are not of paramount importance for left ventricular hypertrophy in human essential hypertension. It is noteworthy that sex, body size and blood pressure explain 64% of the variance of left ventricular mass and 54% of the variance of mean wall thickness in the present study population. Her- edity33 and various lifestyle factors which have not been assessed, such as physical activity,43 alcohol consumption44 and smoking,45 may have contrib- uted to at least part of the remaining unexplained variance, but we found no relationships of cardiac structure with age, 24-h urinary sodium excretion, plasma noradrenaline and previous anti-hypertensive treatment. Results in the literature on the relationships of left ventricular structure with components of the renin-angiotensin-system have been conflicting. Various populations have been studied, ie, young healthy adults,20 employed sub- jects16 or middle-aged men in a general population,15 children and adolescents18 or adults with essential hypertension,13,14,17,19,22 and patients with renovascular hypertension.21 Most authors did not observe a significant relationship of left ventricular Figure 3 Partial regression plots of peak velocity (upper panel) and flow deceleration (lower panel) of the left ventricular early mass (index) or wall thickness, or both, with plasma filling wave (E), against plasma atrial natriuretic peptide, after renin activity.13–15,17,18,21 Left ventricular mass adjustment for the linear effects of age. Adjustments are (index) correlated significantly though weakly (0.30 expressed as standardized residual values. Ͻ r Ͻ 0.35) with plasma renin activity in one report in young healthy adults20 and another in hyperten- analysis (Table 2), but the effect was only inde- sives19 but independence of possible confounding pendent of significant confounders, ie, age and mean factors was not assessed. By contrast, stepwise mul- blood pressure, for aldosterone (Table 3). tiple logistic regression revealed that left ventricular hypertrophy was even associated with low renin levels in patients with sustained hypertension.16 Discussion Whereas left ventricular mass (index) was not The main findings of the present study are: (1) that related to angiotensin II,15,17 significant correlation components of the renin-angiotensin-aldosterone- coefficients have been observed between angioten- system and its physiologic antagonist atrial natriur- sin II and wall thickness or left ventricular mass etic peptide are more closely related to echocardio- (0.22 Ͻ r Ͻ 0.46), which remained significant after

Table 3 Determinants of left ventricular inﬂow characteristics: partial regression coefﬁcients from stepwise multiple regression analysis

2 Epv (m/s) Edec (m/s )Apv (m/s) EA ratio

Age (yr) −0.00532+ −0.0694+ +0.0030** −0.0189+ Sex (m=1;f=2) NS NS +0.0453* NS Mean blood pressure (mm Hg) NS NS NS −0.00448* Heart rate (beats/min) NS NS +0.0030* NS Aldosterone (nmol/L) −0.180** −2.42** NS −0.391** Atrial natriuretic peptide (log pmol/L) +0.126** +1.22* +0.0773* NS Intercept +0.73 +6.84 −0.0055 +2.77 R2 0.33+ 0.36+ 0.39+ 0.53+

*P р0.05; **P р 0.01; +P р0.001; NS: not significant. Other variables offered for selection: plasma renin activity; plasma angiotensin II; plasma noradrenaline; height; weight; left ventricular mass; mean wall thickness; left ventricular internal diameter; 24-h urinary Na excretion; previous medication; duration of hypertension. Abbreviations: see Table 2. Hormones and ventricular diastolic function RH Fagard et al 200 adjustment for confounders such as age, sex, body inflow pattern may occur, but this is likely to be con- size, blood pressure and/or urinary sodium fined to patients with severe hypertension and excretion.17,20,22 Aldosterone was usually not a evolving heart failure. We do not believe that determinant of left ventricular structure.17,18,20,22 pseudo-normalisation confounded our results Hartford et al15 reported a weak association between because (1) patients with a history of heart failure left ventricular mass and aldosterone before (r = were excluded, (2) systolic function, assessed by the 0.26) but not after adjustment for blood pressure. fractional shortening of the left ventricular internal Only Bauwens et al19 found a relatively strong diameter, was normal, and (3) there was no evidence relationship between left ventricular mass index and of reversal of the transmitral inflow pattern in the aldosterone (r = 0.56) in hypertensives, which patients with the highest blood pressure. It is there- remained significant when blood pressure was taken fore likely that the observed abnormalities during into account. More recently, Rossi et al23 reported early left ventricular filling, after consideration of that left ventricular mass and wall thickness were other possible determinants such as age, heart rate, higher in patients with primary aldosteronism com- left ventricular mass and indices of afterload (blood pared with those with essential hypertension. pressure) and preload (left ventricular internal Another approach has been to study the regression diameter),30 are indicative of impaired diastolic of left ventricular hypertrophy by drugs with differ- function, myocardial stiffness, increased collagen ent influences on the renin-angiotensin-system. A content and fibrosis.46 To the best of our knowledge, recent meta-analysis which was restricted to ran- relationships between early transmitral flow charac- domised comparative studies revealed that there is teristics, and, respectively, atrial natriuretic peptide as yet no convincing evidence that anti-hypertensive and effector hormones of the renin-angiotensin-sys- agents which suppress the renin-angiotensin-system tem, have not been reported in the literature, but the reduce left ventricular mass to a greater extent than results are in agreement with the changes of left ven- other first line drugs to which they were directly tricular diastolic function which have been observed compared.29 in primary aldosteronism.23 An inverse relationship of left ventricular mass Peak flow velocity during atrial filling was not sig- and wall thickness with plasma atrial natriuretic nificantly related to components of the renin-angio- peptide might have been expected based on its anti- tensin-system but positively to atrial natriuretic pep- growth and antimitogenic properties, but this was tide. Whereas there are no data on the renin- ` not observed in the present study or in published angiotensin-system, Dessı-Fulgheri et al25 found a reports in normal subjects,24 hypertensive similar relationship of atrial natriuretic peptide with patients,22,24,27 both combined,25 or in patients with atrial peak velocity, whereas Ganau et al24 described congestive heart failure due to diastolic dysfunc- concordant results by use of nuclear angiography, ` tion.26 Dessı-Fulgheri et al25 and Schroeder et al22 suggesting that atrial natriuretic peptide is associa- even observed a positive association between left ted with diastolic dysfunction. This apparent con- ventricular mass index and atrial natriuretic pep- flict with the findings during early inflow can most tide, which remained significant after controlling for likely be explained by the fact that atrial involve- significant confounders in the latter study.22 This ment in hypertensive heart disease not only stimu- finding may be ascribed to atrial natriuretic peptide lates the secretion of atrial natriuretic peptide but secretion by the hypertrophied ventricle, which is also affects the atrial filling characteristics. normally confined to the atrium. The present findings could be criticized because Similar to the results in normal subjects,42 peak they merely represent cross-sectional associations velocity of the early filling wave decreased with age, between aldosterone, atrial natriuretic peptide and whereas the atrial peak velocity increased with age. indices of diastolic function in a particular period After controlling for age (and sex), peak velocity and in the patients’ life, whereas left ventricular deceleration of the early filling wave were inversely remodeling is a long-term process. However, the related to circulating aldosterone and positively to results are plausible and consistent with results from atrial natriuretic peptide, suggesting that aldos- animal experiments in which long-term infusion of terone is associated with reduced diastolic function aldosterone led to myocardial fibrosis.6–9 In addition, and atrial natriuretic peptide with improved dias- the human heart seems to possess the cellular tolic function. It is likely that the echocardiographic machinery for direct aldosterone action, because of measurements can be considered as a surrogate for the coexpression of the mineralo-corticoid receptor myocardial fibrosis. Indeed, Sugihara et al31 demon- and the required protecting enzyme 11 ␤-hydroxys- strated that the increment of left ventricular volume teroid dehydrogenase II.47 It should be noted, how- during the rapid filling phase, assessed by use of M- ever, that the observation that aldosterone increases mode echocardiography, correlated inversely with the rate of synthesis of collagen by cultured rat car- myocardial interstitial fibrosis, measured on right diac fibroblast10 has not been confirmed by others.11 ventricular endomyocardial biopsies in six normo- As for atrial natriuretic peptide it has recently been tensive and 18 hypertensive subjects; this relation- shown that cardiac fibroblasts are capable of ship was independent of wall thickness. Though expressing atrial natriuretic peptide receptor genes48 these authors did not use Doppler velocimetry, the and that atrial natriuretic peptide inhibits cardiac rate of left ventricular filling volume corresponds fibroblast proliferation.12 Finally the cross-sectional with Doppler measurements during early filling. A approach is more likely to weaken associations, so possible limitation of transmitral Doppler velocime- that significant correlations may represent true and try is that ‘pseudo-normalisation’ of the abnormal possibly causal relationships; on the other hand, Hormones and ventricular diastolic function RH Fagard et al 201 nonsignificant relationships may not allow to refute 16 Hammond IW, et al. The prevalence and correlates of the existence of true pathophysiological influences echocardiographic left ventricular hypertrophy among of biochemical variables on cardiac structure and employed patients with uncomplicated hypertension. function. J Am Coll Cardiol 1986; 7: 639–650. 17 Schmieder RE et al. Does the renin-angiotensin-aldosterone system modify cardiac structure and function Acknowledgements in essential hypertension? Am J Med 1988; 84 (Suppl 3A): 136–139. The authors gratefully acknowledge the secretarial 18 Daniels SD, Meyer RA, Loggie JMH. Determinants of and technical assistance of N Ausseloos, J Delsu- cardiac involvement in children and adolescents with pehe, L Lommelen, Y Piccart and J Romont. R essential hypertension. Circulation 1990; 82: 1243– Fagard is holder of the Prof A Amery Chair in 1248. Hypertension Research, founded by MSD (Belgium). 19 Bauwens FR et al. Influence of the arterial blood pressure and nonhemodynamic factors on left ventricular hypertrophy in moderate essential hypertension. Am J References Cardiol 1991; 68: 925–929. 20 Harrap SB et al. Plasma angiotensin II, predisposition 1 Morgan HE, Barker KM. Cardiac hypertrophy. Mechan- to hypertension and left ventricular size in healthy ical, neural and endocrine dependence. Circulation young adults. Circulation 1996; 93: 1148–1154. 1991; 83: 13–25. 21 Yoshitomi Y et al. Comparison of changes in cardiac 2 Schelling P, Fischer H, Ganten D. Angiotensin and cell structure after treatment in secondary hypertension. growth: a link to cardiovascular hypertrophy. J Hyper- Hypertension 1996; 27 (Part 1): 319–323. tens 1991; 9: 3–15. 22 Schroeder AP et al. Influence of humoral and neuro- 3 Weber KT, Brilla CG. Pathological hypertrophy and hormonal factors on cardiovascular hypertrophy in cardiac interstitium. Fibrosis and renin-angiotensin- untreated essential hypertensives. Am J Hypertens aldosterone system. Circulation 1991; 83: 1849–1865. 4 Weber KT, Sun Y, Guarda E. Nonclassic actions of 1996; 9: 207–215. angiotensin II and aldosterone in nonclassic target 23 Rossi GP et al. Changes in left ventricular anatomy and tissue (the heart): Relevance to hypertensive heart dis- function in hypertension and primary aldosteronism. ease. In: Laragh JH, Brenner BM (eds). Hypertension: Hypertension 1996; 27: 1039–1045. Pathophysiology, Diagnosis and Management, 2nd 24 Ganau A et al. Plasma atrial natriuretic factor in essen- Edition. Raven Press Ltd: New York, USA, 1995, tial hypertension: relation to cardiac size, function and pp 2203–2223. systemic hemodynamics. J Am Coll Cardiol 1989; 14: 715–724. 5 Lewicki JA, Protter AA. Physiological studies of the ` natriuretic peptide family. In: Laragh JH, Brenner BM 25 Dessı-Fulgheri P et al. High levels of plasma atrial (eds). Hypertension: Pathophysiology, Diagnosis and natriuretic factor and impaired left ventricular dias- Management, 2nd Edition. Raven Press Ltd: New York, tolic function in hypertensives without left ventricular USA, 1995, pp 1029–1053. hypertrophy. J Hypertens 1992; 10: 161–165. 6 Brilla CG et al. Remodeling of the rat right and left 26 Lang CC et al. Increased plasma levels of brain natriur- ventricles in experimental hypertension. Circ Res etic peptide in patients with isolated diastolic dys- 1990; 67: 1355–1364. function. Am Heart J 1994; 127: 1635–1636. 7 Brilla CG, Weber KT. Reactive and separative myocar- 27 Kohno M et al. Brain natriuretic peptide as a marker dial fibrosis in arterial hypertension in the rat. Car- for hypertensive left ventricular hypertrophy: changes diovasc Res 1992; 26: 671–677. during 1-year antihypertensive therapy with angioten- 8 Robert V et al. Increased cardiac types I and II Collagen sin-converting enzyme inhibitor. Am J Med 1995; 98: 257–265. mRNAs in aldosterone-salt hypertension. Hyperten- ¨ sion 1994; 24: 30–36. 28 Dahlof B, Pennert K, Hansson L. Reversal of left ven- 9 Young M, Fullerton M, Dilley R, Funder J. Mineralo- tricular hypertrophy in hypertensive patients. A meta- corticoids, Hypertension and Cardiac Fibrosis. J Clin analysis of 109 treatment studies. Am J Hypertens Invest 1994; 93: 2578–2583. 1992; 5: 95–110. 10 Brilla CG, Zhou G, Weber K, Matsubara L. Collagen 29 Fagard RH. Reversibility of left ventricular hypertro- metabolism in cultured adult rat cardiac fibroblasts: phy by antihypertensive drugs. Neth J Med 1995; 47: response to angiotensin II and aldosterone. J Mol Cell 173–179. Cardiol 1994; 26: 809–820. 30 Harizi RC, Bianco JA, Alpert JS. Diastolic function of 11 Fullerton MJ, Funder JW. Aldosterone and cardiac the heart in clinical cardiology. Arch Int Med 1988; fibrosis: in vitro studies. Cardiovasc Res 1994; 28: 148: 99–109. 1863–1867. 31 Sugihara N et al. Diastolic dysfunction and its relation 12 Fujisaki H et al. Natriuretic peptides inhibit angioten- to myocardial fibrosis in essential hypertension. J Car- sin II-induced proliferation of rat cardiac fibroblasts by diol 1988; 18: 353–361. blocking endothelin-1 gene expression. J Clin Invest 32 Guidelines Subcommittee. 1993 Guidelines for the 1995; 96: 1059–1065. Management of Mild Hypertension: Memorandum 13 Devereux RB, Savage DD, Drayer JIM, Laragh JH. Left from a WHO/ISH meeting. J Hypertens 1993; 11: ventricular hypertrophy and function in high, normal, 905–918. and low-renin forms of essential hypertension. Hyper- 33 Bielen E, Fagard R, Amery A. The inheritance of left tension 1982; 4: 524–531. ventricular structure and function assessed by imaging 14 Rowlands DB et al. Assessment of left ventricular mass and Doppler echocardiography. Am Heart J 1991; 121: and its response to antihypertensive treatment. Lancet 1743–1749. 1982; 1: 467–470. 34 Sahn DJ, De Maria A, Kisslo J, Weyman A. Recommen- 15 Hartford M et al. Left ventricular mass in middle-aged dations regarding quantitation in M-mode echocardio- men: Relationship to blood pressure, sympathetic ner- graphy: results of a survey of echocardiographic vous activity, hormonal and metabolic factors. Clin measurement. Circulation 1978; 58: 1072–1083. Exp Hyper- Theory and Practice 1983; A5: 1429–1451. 35 Devereux RB et al. Echocardiographic assessment of Hormones and ventricular diastolic function RH Fagard et al 202 left ventricular hypertrophy: comparison to necropsy Linker DT (eds). Cardiac Ultrasound. Churchill Liv- findings. Am J Cardiol 1986; 57: 450–458. ingstone: Edinburgh, 1993, pp 141–153. 36 Fagard R, Bielen E, Lijnen P, Amery A. Cardiac vari- 43 Fagard R. Impact of different sports and training on ables and blood pressure as determinants of left ven- cardiac structure and function. Cardiol Clin 1992; 10: tricular inflow velocities. J Hum Hypertens 1993; 7: 241–256. 7–12. 44 Manolio TA et al. Relation of alcohol intake to left ven- 37 Fyhrquist F, Puutula L. Faster radioimmunoassay of tricular mass: The Framingham Study. J Am Coll Car- angiotensin I at 37°C. Clin Chem 1978; 24: 115–118. diol 1991; 17: 717–721. 38 Lijnen P, Amery A, Fagard R, Katz F. Radioimmunoas- 45 Liebson PR et al. Echocardiographic correlates of left say of angiotensin II in unextracted plasma. Clin Chim ventricular structure among 844 mildly hypertensive Acta 1978; 88: 403–412. men and women in the Treatment of Mild Hyperten- 39 Lijnen P, Amery A, Fagard R, Corvol P. Direct radioim- sion Study (TOMHS). Circulation 1993; 87: 476–486. munoassay of plasma aldosterone in normal subjects. 46 Doering CW et al. Collagen network remodelling and Clin Chim Acta 1978; 84: 305–314. diastolic stiffness of the rat left ventricle with pressure 40 Lijnen P et al. Plasma atrial natriuretic peptide and overload hypertrophy. Cardiovasc Res 1988; 22: 686– cyclic nucleotide levels before and after a marathon. J 695. Appl Physiol 1987; 63: 1180–1184. ` 47 LombesMet al. Prerequisite for cardiac aldosterone 41 Bauersfeld W, Ratge D, Knoll E, Wisser H. Bestimmung action. Mineralocorticoid receptor and 11 ␤-HSD in der Catecholamine in Plasma mit HPLC und ampero- the human heart. Circulation 1995; 92: 175–182. metrischer Detektion. Fresenius Z Anal Chem 1984; 317: 679–680. 48 Lin X et al. Gene expression of natriuretic peptide 42 Rinaldi JP, Diebold B. Normal intracardiac Doppler receptors in myocardial cells. Circ Res 1995; 77: flows. In: Roelandt JRTC, Sutherland GR, Iliceto S, 750–758.