High Prevalence of Concentric Remodeling in Elderly Individuals With Isolated Systolic Hypertension From a Population Survey
Echocardiographic determination of left ventricular mass index (LVMI) is shown to be valuable in the assessment of cardiovascular risk. Determination of left ventricular geometry, including concentric remodeling, provides additional prognostic information. In isolated systolic hypertension (ISH), the few echocardiographic studies available show an increased LVMI, but criteria and patient populations differ. No comparison with diastolic hypertension (DH) has been made, nor has left ventricular geometry (with concentric remodeling) been evaluated. We compared both LVMI and left ventricular geometry of newly diagnosed ISH subjects with normotensive and DH subjects, all previously untreated and from the same population. The echocardiographic LVMI of 97 previously untreated ISH subjects (4× systolic pressure ≥160 mm Hg, diastolic pressure <95 mm Hg) was clearly elevated compared with values in age- and sex-matched normotensive subjects (98 and 71 g/m2, respectively; P<.001). The geometric pattern was abnormal in most ISH subjects, with a high prevalence (43%) of concentric remodeling. Both LVMI and left ventricular geometry of ISH subjects did not differ significantly from values in DH subjects (LVMI, 92 g/m2; concentric remodeling, 56%). Sex differences in LV geometry in ISH were present only with the Framingham criteria, not with the Koren criteria. This study shows a high prevalence of concentric remodeling in elderly individuals with previously untreated ISH. The increase of LVMI and abnormality in left ventricular geometry are comparable with those in DH subjects, further defining the place of ISH as a cardiovascular risk factor in the elderly. Whether there are sex differences in cardiac adaptation in ISH and whether the geometric classification can be used to adjust treatment remain to be investigated.
Isolated systolic hypertension (ISH) has long been known as an important cardiovascular risk factor,1 2 3 4 and the results of the Systolic Hypertension in the Elderly Program (SHEP) have shown that its treatment can lead to a favorable outcome.5 The high prevalence of ISH in the elderly represents a major public health issue.6 Decisions regarding its management and treatment can be supported by knowledge about the presence of cardiovascular end-organ damage in these individuals. An accepted method for assessing this damage in hypertensive individuals is measurement of echocardiographic LVMI. In DH, the occurrence of increased LVMI, with or without meeting the criteria for LVH, is a strong and independent risk factor for cardiovascular events in different populations, including the elderly.7 8 9 10
Koren et al11 also showed that LV geometry has prognostic significance. The relationship of cardiovascular risk to the pattern of geometry (normal<concentric remodeling<eccentric LVH<concentric LVH) was recently confirmed in hypertensive patients in a study by Verdecchia et al.12 In contrast to the original article by Koren et al, Verdecchia et al included interventricular septal thickness in the calculation of LV geometry instead of only posterior wall thickness. This may alter the results because, as they have described, isolated septal thickening might occur in hypertension.13 Another point of discussion is whether sex-dependent or sex-independent criteria for LVH should be used; this obviously influences any analysis of differences between men and women in LV adaptation in hypertension.14
So far, few studies have addressed LVMI in the elderly with ISH. Comparison of the results of these studies is hampered by differences in methodology and participant recruitment: the largest study (Cardiovascular Health Study) is unclear in the indexation of LV mass, whereas others are done in selected groups of individuals or use different criteria.14 15 16 No comparisons with elderly DH patients are presented, and none of these studies has determined LV geometry. The objective of the present study was to determine the effect of previously untreated ISH on LV mass and geometry in elderly individuals selected from a population survey. The results are compared with an age- and sex-matched normotensive control group and with DH subjects from the same population survey.
A population screening program was done in a rural municipality (Achtkarspelen) in the Netherlands. All inhabitants between 60 and 75 years of age were invited to participate. After subjects had rested 5 minutes, sitting BP was measured by trained volunteers using a sphygmomanometer. If BP was elevated (for systolic BP, ≥160 mm Hg), the BP measurement was repeated by a physician. If BP remained elevated, the subject was invited for a third and fourth measurement on two separate occasions. Subjects were defined as having ISH if untreated systolic BP was greater than or equal to 160 mm Hg at each measurement, with diastolic BP less than 95 mm Hg on at least the last two occasions, and average diastolic BP less than 95 mm Hg. All individuals fulfilling these criteria were eligible for investigation of echocardiographic end-organ damage. Exclusion criteria for this study were recent myocardial infarction (<3 months), cerebrovascular accident (<1 year), symptomatic congestive heart failure, hemodynamically significant valvular heart disease or cardiac arrhythmia (other than atrial fibrillation), or any other disabling medical condition hampering subject participation. Subjects were also excluded if both apical and parasternal echocardiographic views were inadequate for obtaining measurements. The general physician was consulted about subject eligibility for further investigation. The study was approved by the Medical Ethics Committee of the Groningen University Hospital, and written informed consent was obtained from all subjects.
A group of normotensive control subjects was selected from the same program (BP at first measurement <160/95 mm Hg, and no known hypertension). They were matched to the ISH subjects for age, sex, and a functional classification (Duke activity status index).17
Results of ISH subjects were compared with those of DH subjects. This group consisted of all DH subjects found in the same population survey, with DH defined as untreated average diastolic BP greater than or equal to 95 mm Hg (and <115 mm Hg), measured on three separate occasions. Exclusion criteria were the same as for ISH subjects, with addition of diabetes mellitus (due to a planned intervention trial).
All ISH and DH subjects and normotensive control subjects underwent 24-hour ambulatory BP measurement (SpaceLabs 90207, SpaceLabs Inc). Measurements were done every 30 minutes during the day (7 am to 11 pm) and every 60 minutes during the night (11 pm to 7 am); data were analyzed without data editing using time-weighed means. Histories of smoking, diabetes, and myocardial infarction were obtained. Height and weight were measured, and body mass index was calculated. The fasting serum lipid profile, including high-density lipoprotein cholesterol and triglycerides, was determined with conventional assay methods. Fasting serum glucose was also determined; subjects with plasma glucose greater than 7.8 mmol/L (World Health Organization criterion) were considered to be diabetic, in addition to those already known to have or be on treatment for diabetes mellitus.
Echocardiographic Examination and LV Mass Calculations
Echocardiographic dimensions were measured with a 2- to 2.5-MHz transducer (128 XP-5, Acuson Corp) according to the Penn convention with subjects in the left decubitus position. Measurements of LV internal end-diastolic (EDD) and end-systolic diameters and of interventricular septal (IVS) and posterior wall (PWT) thicknesses were obtained from a standard parasternal long-axis view. Measurements were made from two-dimensional images perpendicular to the long axis of the left ventricle. The left atrial dimension was measured in both the parasternal and apical views. All dimensions are averages of three end-expiratory measurements in sinus rhythm (five in case of atrial fibrillation). Measurements were performed on-line by the same echocardiographer and were videotaped. Intraobserver variability of the echocardiographer (W.F.H.), measured before the start of the study, was within a 5% error margin. LV mass was calculated according to the formula as corrected by Devereux and Reichek18 19 —LVMI=1.04·[(IVS+PWT+EDD)3−EDD3]−13.6—and indexed for body surface area. In the geometric analysis of Koren et al,11 LV geometry is subdivided using the presence or absence of LVH (defined as LVMI ≥125 g/m2) and of increased RWT (RWT=2×PWT/EDD; cutoff value, ≥0.45). A normal RWT and no LVH is defined as normal geometry; LVH is subdivided into concentric LVH when RWT is increased and eccentric LVH when RWT is normal. The pattern of increased RWT without LVH is called concentric remodeling. In ISH, the geometric classification is also analyzed by sex to investigate possible differences in cardiac adaptation.
For discussion purposes, we performed two extra calculations. First, we repeated the geometric classification using the Framingham criteria, with LVMI indexed for height and cutoff values by sex for LVH (143 g/m for men, and 102 g/m for women).20 Also, RWT was recalculated with the septal wall thickness and with the sum of the posterior and septal walls in the calculation of LV geometry.
Data are presented as mean±SD. Differences among the three groups were analyzed with Pearson χ2 testing in the case of nominal data (Fisher's exact test in the case of low numbers) and with one-way ANOVA in the case of continuous variables, with post hoc corrections for multiple comparisons (Tukey's honestly significant difference). Differences between two groups, as in analysis of sex, were calculated by Student's t test. Odds ratios were calculated for ISH subjects, with the normotensive subjects as case-controls, and are presented with 95% confidence intervals. Differences were considered statistically significant when the two-sided probability values were less than .05 (P<.05).
Of the 2776 inhabitants aged 60 to 74 years, 1812 (65.2%) attended the screening. After four measurements, 150 people met the criteria for ISH (prevalence, 8.3%). Of these, 97 subjects participated in further echocardiographic examination. Reasons for ineligibility were mainly lack of interest by the individuals (n=30) or being considered unsuitable for investigation by the general practitioner consulted (n=12); major exclusion criteria were met by 11 subjects. The results of echocardiographic examinations were compared with those of 41 patients with previously untreated DH and with 97 normotensive control subjects matched to ISH subjects for age and sex.
Baseline Characteristics and BP Measurements
Table 1⇓ presents the baseline characteristics of the three groups, including the other cardiovascular risk factors. Age and sex of ISH and control subjects were equal because of matching. The DH subjects, although recruited from the same age category of 60 to 74 years, were slightly, but significantly, younger than the ISH subjects (and control subjects). Sex distribution of DH subjects did not differ from that of ISH and control subjects. Body mass index of DH subjects was significantly higher than that of control subjects but not that of ISH subjects. Body surface area or length, which is used for indexation of LV mass, did not differ significantly among the three groups. Some significant differences could be observed in the other cardiovascular risk factors between the two hypertensive groups and the normotensive control group: both hypertensive groups contained more smokers, and the ratio of total to high-density lipoprotein cholesterol was higher in the control group (because of lower high-density lipoprotein cholesterol).
Table 2⇓ shows the results of office and ambulatory BP measurements. Most of the differences in office BP can be explained by the different entrance criteria of the three groups; these differences were confirmed in patterns of 24-hour ambulatory measurements. All values of ISH subjects were significantly higher than those of normotensive control subjects, including diastolic BP. Diastolic BP values of DH subjects were higher than those of ISH and control subjects. Systolic BP of the ISH subjects did not differ from that of DH subjects, for both office and ambulatory measurements. The difference between daytime and nighttime measurements (the “dipping” phenomenon) was the same in all three groups.
The results of echocardiographic measurements are presented in Table 3⇓. One ISH subject and three control subjects had inadequate parasternal views for LV mass measurement; therefore, the results are shown of 96 ISH, 94 control, and 50 DH subjects. Comparison of ISH subjects with normotensive control subjects showed that LVMI was significantly increased in ISH subjects because of thickening of both the interventricular septum and posterior wall and because of an increase in end-diastolic dimension. The relative difference in wall thickness appeared to be larger than the increase in end-diastolic dimension, as shown by the significantly increased RWT of ISH subjects. The percentage of LVH (defined as LVMI ≥125 g/m2) in ISH subjects was only slightly and not significantly higher than that of normotensive control subjects. The left atrial dimension was significantly increased in ISH compared with control subjects in the parasternal view only. All the results of LV mass measurements were comparable in ISH and DH subjects, with LVMI, RWT, and both wall and end-diastolic dimensions showing no significant differences. The results of the LV geometric classification according to Koren are depicted in Fig 1⇓. Overall, LV geometry of ISH subjects clearly differed from that in normotensive control subjects (P<.001). Most of the normotensive control subjects (86%) had a normal geometry, versus 48% of ISH subjects. Total LVH prevalence in ISH subjects was 9% (control subjects, 3%), without an apparent difference between concentric and eccentric LVH. Concentric remodeling was found in 43% of ISH subjects versus 12% of control subjects; expressed as an odds ratio, ISH subjects were 6.5 times more likely (95% confidence interval, 3.0-13.8; P<.001) to have this pattern than control subjects. Almost half (47%) of ISH subjects who might have been considered “normal” because of an absence of LVH had this abnormal pattern of increased wall thickness.
No significant difference was found in LV geometric pattern between ISH and DH subjects (Pearson χ2, P=.423). The percentage of LVH in both groups was 9%; concentric remodeling in DH subjects was present in 56%.
Sex Differences and Influence of Different Criteria and Formulas
With the Koren criteria, analysis of LV geometry by sex in ISH subjects showed no significant differences between men and women (Fig 2⇓, top). Eccentric LVH appeared to be more frequent in men with ISH; when LVH subjects were analyzed separately, this difference was of borderline significance (P=.058). When the sex-dependent criteria of the Framingham Heart Study were applied to this population, the results were remarkably different and did reach overall statistical significance (P<.05; Fig 2⇓, bottom). Female ISH subjects more often showed an abnormal geometric pattern: the overall percentage of LVH was 43% with the Framingham criteria versus 6% with the original Koren criteria. These different results appeared to be solely due to a different classification of women; the results of men were comparable for the two criteria.
Concerning asymmetry of LVH, we did not often find isolated septal wall thickening in either ISH or DH subjects and normotensive control subjects. Only two ISH subjects had a ratio of greater than or equal to 1.3 of one wall thickness to the opposite side13 ; however, in both cases, this was increased thickness of the posterior wall instead of increased septal thickening. The influence of LV wall thickness being used in the calculation of RWT, and the resulting LV geometric classification, was only minor. The percentage of increased RWT varied little when the original formula was applied with posterior wall thickness (rate, 34% for all subjects), septal thickness (rate, 32%), or the sum of both walls in the formula (30%).
This study shows that the echocardiographic LV mass and geometric pattern in individuals with previously untreated ISH are clearly abnormal, comparable with results of individuals with DH and significantly different from results of normotensive control subjects. The pattern of abnormality was mainly an increase of wall thickness, apart from enlargement of the ventricles. Results of the geometric classification showed that the majority of ISH subjects had an abnormal LV geometry, with a moderate prevalence of LVH but a high prevalence of concentric remodeling. This abnormal pattern of increased wall thickness was found in almost half of the ISH subjects, who were otherwise perhaps considered normal because of the absence of LVH. We therefore conclude that these individuals with ISH from a general population clearly show an increased prevalence of cardiac end-organ damage, to a level comparable with that in individuals with DH. The importance of this observation is that all subjects in this population-based study were previously untreated, which avoids any interference of medication on the results. With the proven attributable cardiovascular risk of increased LVMI and abnormal LV geometry, this further defines the cardiovascular risk of ISH in a general population.
Interestingly, the results in ISH and DH subjects were comparable, despite the fact that the underlying pathophysiological vascular changes are almost opposed in ISH and DH: In ISH, distensibility of larger arteries is increased, with little or no change in peripheral resistance, whereas in DH, peripheral resistance is clearly increased.20 In the absence of (invasive) hemodynamic measurements, no clear explanation can be given except that hemodynamic load, as derived from values of LV mass and geometry, apparently is equal in these subject groups.
Studies on echocardiography in ISH are rather scarce, often including only small numbers of ISH individuals with varying selection criteria.20 21 The size of the present study is second only to the Cardiovascular Health Study and is slightly larger than the SHEP study and the Framingham ISH report.14 15 16 All of these studies have shown a higher LVMI in ISH individuals compared with normotensive control subjects. However, subject selection and criteria used differed among studies, hampering the comparison of results, and none of these studies included a comparable group of individuals with DH. In the Cardiovascular Health Study, only the results of unindexed LV mass were reported, without separate results on wall thickness and internal dimensions, which does not allow analysis of geometric changes.12 Comparison of the results with the ISH subjects from the Framingham Heart Study is complicated by the different indexation of LVMI used (for height instead of for body surface area).14 Such a different indexation may lead to markedly different results, as is shown in our results on sex differences in ISH. The results of LV mass measurements in a subpopulation of the SHEP study are to some extent comparable with those of the present study, although the selection of subjects differs.16 In the SHEP study, only 1% of referred individuals finally participated, including previously treated patients. In our study, almost two thirds (65%) of all detected ISH subjects were included, all of them previously untreated.
In the present study, some possible sources of bias must be discussed. First, although attendance rate was rather high (60%), this study might, as with any screening study, suffer from “voluntary bias”: healthy and health-conscious people may be more likely to attend. To analyze the possibility of differences in BP and health status (number of chronic diseases registered by questionnaire), we conducted a “nonresponder” investigation. Those individuals not attending did suffer from slightly more chronic diseases compared with the attendants as well as a somewhat lower BP, but differences were not significant. Second, of the 150 ISH subjects found in the screening, 97 underwent echocardiographic examination. Although this is a much higher percentage (65%) when compared with some other reports, this can also be a possible source of bias. However, only a minor portion was excluded because of concomitant diseases, among which several were excluded for noncardiovascular reasons; most subjects refused further investigation. Finally, it could be remarked that the inclusion criterion for ISH subjects in the present study was diastolic BP less than 95 mm Hg, as opposed to other (North American) criteria of less than 90 mm Hg. However, this had no influence on the present results: the LVMI of those subjects with inclusion diastolic BP of 90 to 94 mm Hg was 94.7 g/m2 and thus was comparable with the mean value of the total group, and the prevalence of concentric remodeling was 45%, with no subjects with LVH in this group.
Previous studies have suggested a possible difference in cardiac adaptation in ISH with sex: more concentric LVH in women and more eccentric LVH in men.14 This pattern is also observed in elderly individuals with LV pressure overload by aortic stenosis.22 In such analyses, the influence of the use of different criteria is very important. When using the original criteria of Koren, we could not confirm such a difference, although there was a trend toward more eccentric LVH in men. However, when applying the (sex-specific) criteria of the Framingham Heart Study to our subjects, we saw a marked difference in geometry in women, resulting in a significant difference between the geometry of ISH in men and women. This is probably due to the rather large difference in criteria for LVH in men and women—143 versus 102 g/m, respectively. In particular, the prevalence of abnormal patterns in women increased considerably with the application of these criteria.
A further point of discussion is whether only the posterior wall should be used in the calculation of RWT and thus LV geometry (Koren) or the interventricular septal thickness as well (Verdecchia et al13 ). Not only does it seem illogical to define a pattern as “concentric” if only one side of the LV wall is included, more important, Verdecchia et al could not observe any significant difference in cardiovascular risk for concentric remodeling when only posterior wall thickness was included. We found that results were comparable with inclusion of either one or both walls; we did not observe a significant prevalence of isolated septal thickening. In our experience, a large proportion of the elderly shows an angling of the left ventricle toward the echocardiographic transducer, and this is known to be a possible source of overestimation especially of septal thickness with M-mode measurement.23 24 For this reason, we use two-dimensional measurement of wall thickness and lumen diameters in our laboratory, making it possible to do measurements perpendicular to the LV long axis and thus avoid a possible overestimation.
In conclusion, this study is one of the first to show that echocardiographic LVMI and LV geometry in individuals with ISH are comparable with values in individuals with DH. A large number of those individuals perhaps initially considered to be normal because of the absence of LVH show concentric remodeling, which indicates an increased cardiovascular risk for these subjects. These results further define the place of ISH as a cardiovascular risk factor, indicating that the risk of ISH is no less than that of DH in the elderly. It remains to be established which criteria are most predictive and whether results from geometric classification can be used to adjust treatment.
Selected Abbreviations and Acronyms
|ISH||=||isolated systolic hypertension|
|LVH||=||left ventricular hypertrophy|
|LVMI||=||left ventricular mass index|
|RWT||=||relative wall thickness|
This investigation was supported by the “Praeventiefonds,” grant No. 28-2219.
Reprint requests to Wilfred F. Heesen, MD, Department of Cardiology, Academic Hospital Groningen, Hanzeplein 1, PO Box 30.001, 9700 RB Groningen, Netherlands.
- Received August 8, 1996.
- Revision received September 3, 1996.
- Accepted September 3, 1996.
Cubb JD, Borhani NO, Entwisle G, Tung B, Kass E, Schnaper H, Williams W, Berman R. Isolated systolic hypertension in 14 communities. Am J Epidemiol. 1985;121:362-370.
Garland C, Barrett-Connor E, Suarez L, Criqui MH. Isolated systolic hypertension and mortality after age 60 years. Am J Epidemiol. 1983;118:365-376.
Applegate WB. Hypertension in elderly patients (review). Ann Intern Med. 1989;110:901-915.
Levy D, Garrison RJ, Savage DD, Kannell WB, Castelli WP. Left ventricular mass and incidence of coronary heart disease in an elderly cohort. Ann Intern Med. 1989;110:101-107.
Casale PN, Devereux RB, Milner M, Zullo G, Harshfeld GA, Pickering TG, Laragh JH. Value of echocardiographic measurement of left ventricular mass in predicting cardiovascular morbid events in hypertensive men. Ann Intern Med. 1986;105:173-178.
Aronow WS, Ahn C, Kronzon I, Koenigsberg M. Congestive heart failure, coronary events and atherothrombotic brain infarction in elderly blacks and whites with systemic hypertension and with and without echocardiographic and electrocardiographic evidence of left ventricular hypertrophy. Am J Cardiol. 1991;67:295-299.
Koren MJ, Devereux RB, Casale PN, Savage DD, Laragh JH. Relation of left ventricular mass and geometry to morbidity and mortality in uncomplicated essential hypertension. Ann Intern Med. 1991;114:345-352.
Verdecchia P, Schillacci G, Borgioni C, Ciucci A, Battistelli M, Bartoccini C, Santucci A, Santucci C, Reboldi G, Porcellati C. Adverse prognostic significance of concentric remodeling of the left ventricle in hypertensive patients with normal left ventricular mass. J Am Coll Cardiol. 1995;25:871-878.
Verdecchia P, Porcellati C, Zampi I, Schillaci G, Gatteschi C, Battistelli M, Bartoccini C, Borgioni C, Ciucci A. Asymmetric left ventricular remodeling due to isolated septal thickening in patients with systemic hypertension and normal left ventricular masses. Am J Cardiol. 1994;73:247-252.
Devereux RB, Reichek N. Echocardiographic determination of left ventricular mass in men: anatomic validation of the method. Circulation. 1977;55:613-618.
Dart A, Silagy C, Dewar E, Jennings G, McNeil J. Aortic distensibility and left ventricular structure and function in isolated systolic hypertension. Eur Heart J. 1993;14:1465-1470.
Nagano N, Iwatsubo H, Hata T, Mikami H, Ogihara T. Effects of antihypertensive treatment on cardiac hypertrophy and cardiac function in elderly hypertensive patients. J Cardiovasc Pharmacol. 1991;17(suppl 2):S163-S165.
Carroll JD, Carroll EP, Feldman T, Ward DM, Lang RM, McGaughey D, Karp RB. Sex-associated differences in left ventricular function in aortic stenosis of the elderly. Circulation. 1992;86:1099-1107.