|
|
||||||||||
|
J Am Coll Cardiol, 1999; 33:1685-1691 © 1999 by the American College of Cardiology Foundation |

a Klinik und Poliklinik für Innere Medizin II, University of Regensburg, Regensburg, Germany
* Institut für Epidemiologie und Sozialmedizin, University of Münster, Münster, Germany
GSF Forschungszentrum, Institut für Epidemiologie, Munich-Neuherberg, Germany
Manuscript received July 16, 1998; revised manuscript received December 9, 1998, accepted January 20, 1999.
Reprint requests and correspondence: Dr. H. Schunkert, Klinik und Poliklinik für Innere Medizin II, University of Regensburg, D-93042, Regensburg, Germany.
heribert.schunkert{at}klinik.uni-regensburg.de
| Abstract |
|---|
|
|
|---|
The study evaluated the contribution of familial predisposition to the risk of left ventricular hypertrophy (LVH).
BACKGROUND
Left ventricular hypertrophy is a multifactorial condition that serves as an important predictor of cardiovascular mortality. At present it is unclear whether familial predisposition contributes to the manifestation of LVH. Thus, we determined whether siblings of subjects with LVH are at increased risk to present with an elevation of LV mass or an abnormal LV geometry.
METHODS
Echocardiographic and anthropometric measurements were performed in 2,293 individuals who participated in the echocardiographic substudies of population-based MONICA Augsburg surveys. In addition, a total of 319 siblings of survey participants with echocardiographic evidence of LVH were evaluated. The risk of these siblings to present with LVH or abnormal LV geometry was estimated by comparison with 636 subjects matched for gender and age that were selected from the entire echocardiography study base.
RESULTS
Blood pressure, body mass index, age, and gender (i.e., known determinants of LV mass) were comparable in LVH-siblings and the matched comparison group. However, septal and posterior wall thicknesses, relative wall thickness as well as LV mass index were significantly elevated in LVH-siblings (p < 0.001, each) whereas LV dimensions did not differ. Likewise, the prevalence of LVH was raised in LVH-siblings, as was the relative risk of LVH after adjustment for confounders (p < 0.05). More specifically, LVH-siblings displayed increased prevalences of concentric remodeling and concentric LVH (p < 0.05) but not of eccentric LVH.
CONCLUSIONS
Familial predisposition appears to contribute to increased LV wall thickness, to the development of LV hypertrophy and abnormal LV geometry.
| ||||||||||||||||
| Subjects and methods |
|---|
|
|
|---|
Echocardiographic substudies were performed to further evaluate electrocardiographic data suggesting that inheritable factors are involved in the pathogenesis of LVH (12). In 1994, participants of the 1984 to 1985 survey were reinvited for an echocardiographic examination. The invitation was restricted to subjects 52 to 67 years of age, and 646 men and women (response 64%) could be examined (1315). In addition, echocardiographic examinations were offered to all participants of the 1994/1995 survey attending the Augsburg study center. Thereby, another 1,678 men and women in the age range from 25 to 74 years (response 70.6%) could be examined (16).
All participants of the two echocardiographic substudies who displayed LVH by echocardiography (LVM indexed to body height >143 g/m in men and >102 g/m in women; n = 427) were interviewed about the number and gender of their living siblings. Whenever study participants and their siblings agreed, siblings were invited to participate in the examinations applying the same study protocol. A total of 373 male and female siblings were subsequently examined.
From all 2,293 participants of the echocardiographic substudies and all LVH-siblings, data on medical history, medication, and smoking or drinking behavior were obtained by interviews. Subjects body height and weight were measured in light clothing and without wearing shoes, and the body mass index (BMI) was computed as weight in kilograms divided by height in meters squared (kg/m2). Resting blood pressure was measured before and after echocardiography using a standard mercury sphygmomanometer. Blood pressure was read three times at the right arm, and the mean of all three measurements was used in this study. Hypertension was defined as blood pressure values equal or above systolic 160 mm Hg and/or diastolic 95 mm Hg, or as present in subjects chronically taking antihypertensive medication.
Echocardiographic measurements.
Two-dimensional guided M-mode echocardiograms were obtained from each subject by one of three expert sonographers with a Sonos 1500 (Hewlett-Packard, Andover, Massachusetts). The M-mode tracings were recorded on strip chart paper at 50 mm/s, given a code number and analyzed without knowledge on whether the echocardiogram was from a LVH-sibling or a control subject. Only tracings that demonstrated optimal visualization of left ventricular interferences were used. Measurements of wall thicknesses and ventricular diameters for the calculation of LVM were made in random order on coded strip charts without knowledge of whether the recording was from a sibling or a matched subject. Left ventricular internal end-diastolic diameter (EDD), septal wall thickness (SWT), and posterior wall thickness (PWT) were measured according to the Penn convention just below the tip of the mitral valve as recommended by the American Society of Echocardiography (17). Left ventricular mass (LVM) was calculated according to the formula of Devereux et al. (18):
![]() |
The LVM indexed to body size (LVMI) was obtained in three different ways: it was either divided by body height in meters (19), by the allometric signal height2.7 (20), or by body surface area in meters squared (19). The following LVH criteria for each indexation method were used: LVM indexed to height: >143 g/m in men, >102 g/m in women (19); LVM indexed to body surface area: >131 g/m2 in men, >100 g/m2 in women (19); LVM indexed to height2.7: >50 g/m2.7 in men, >47 g/m2.7 in women (20). Patterns of left ventricular geometry were defined as proposed by Ganau et al. (21)that is, as normal geometry when LVMI was normal and relative wall thickness (RWT = 2 PWT/EDD) <0.45; as left ventricular remodeling when normal LVMI was combined with RWT >0.45; as concentric LVH when left ventricular hypertrophy occurred with a RWT >0.45; and as eccentric LVH when left ventricular hypertrophy and a RWT <0.45 were combined.
Statistical analyses. To investigate the influence of familial predisposition on risk of LVH we used a method proposed by Weis et al. (22,23). We compared each sibling with two control subjects from the population-based echocardiographic substudies of the MONICA Augsburg project. For this purpose, each LVH-sibling was randomly matched with two individuals of the same gender and five-years age group. We thus avoided a direct comparison of sibling pairs (i.e., correlation analysis), which might be criticized for the fact that various factors can introduce high correlations between siblings rather indirectly (with respect to the hearte.g., body size) or may be explained by factors beyond inheritance ("shared environment" in a large sense). By distinction, we examined in our study design how the "exposure" of individuals to familial predisposition, defined as having a sibling diagnosed with LVH, may relate to left ventricular mass and hypertrophy aside from the amount contributed by age, gender, body mass, and blood pressure. The latter factors were controlled by design either through appropriate matching (age, gender) or through statistical adjustment (BMI and systolic blood pressure).
Echocardiographic tracings that demonstrated optimal visualization of left ventricular inferences and a complete set of analytic variables were obtained from 319 LVH-siblings (85% of the total of 373 examined) and from 1,923 individuals from the survey samples (84% of the total of 2,293). Excluded subjects were significantly older, more obese and more often men (each p < 0.01). The analyses in this report are based on a comparison 319 LVH-siblings with 636 subjects who were randomly selected after successful matching by gender and age group.
The description of baseline data compares the mean values of age, blood pressure, and body size between LVH-siblings and the comparison group. The statistical significance of differences in continuous variables was assessed by two-sided t tests for unpaired groups and for categorized variables by chi-square tests. Mean values of echocardiographic measurements were compared after adjustment for age, gender, systolic blood pressure, and BMI by means of multiple linear regression modeling (PROC GLM of the SAS statistical software, Version 6.11). Furthermore, the prevalences of LVH using indexations for body height, height2.7, and body surface area were assessed. Multiple logistic regression was applied to estimate the ratio of the LVH odds in LVH-siblings and their comparisons controlling for age, gender, systolic blood pressure, and BMI. The 95% confidence intervals are reported as a measure of the precision of the odds ratio estimates. Finally, the geometric patterns of LVH were analyzed in both siblings and controls. The odds ratio for the occurrence of a concentric pattern (relative wall thickness >0.45) was estimated by logistic regression controlling for covariates.
| Results |
|---|
|
|
|---|
|
|
|
|
|
| Discussion |
|---|
|
|
|---|
LVH a familial risk factor. After adjustment for age, blood pressure, and BMI, the risk elevation related to a familial disposition of LVH was 1.4 to 1.7. This proportion compares favorably to that calculated in recent twin studies for the genetic component of LVM variability (3,68). It needs to be pointed out, however, that the present data do not allow one to distinguish whether the elevated risk of LVH is related to the fact that the siblings shared genes, environment, or both (3,58,26). Nevertheless, the consideration of bothtwin studies and the present investigation on siblings of subjects with LVHallows the hypothesis that LVH is modulated, in part, by an inherited component. This view is strengthened by a small but significant effect of heredity on LVM observed in adult siblings of the Framingham Heart Study (9). Furthermore, the present analysis and two previous twin studies concur in the observation that wall thickness or LVM (I) rather than left ventricular dimension is affected by a familial (and presumably genetic) predisposition (7,8). Consequently, we observed a familial risk for the manifestation of concentric LVH and concentric left ventricular remodeling, both being characterized by increased wall thickness, but not eccentric LVH, which is predominantly related to an increase in left ventricular dimension.
Interaction with other determinants of LVH. This study was carried out on population-based samples to avoid a preselection of LVH cases, which occurs in the clinical setting. It should be emphasized, therefore, that about one-third of the individuals with LVH were normotensive and with normal BMI and, thus, without traditional risk factors for LVH. This distribution resembles observations from other population-based surveys such as the Framingham Heart Study (12,27). We thus conducted subanalyses that revealed a familial component of increased RWT in LVH-siblings irrespective of blood pressure or body mass status. Although an observational study cannot entirely rule out secondary effects on LV mass, consistent findings after adjustment as well as stratification by blood pressure and weight status suggest that the familial risk of LVH, like the variability of left ventricular mass in twins (3), is not secondary to a familial predisposition for hypertension or obesity.
Implications for molecular genetics. Recent molecular genetic investigations suggested that the risk to present with LVH may be affected by alterations of specific candidate genes, such as the angiotensin-converting enzyme gene (12). Although a series of studies reproduced this observation, others were not able to confirm the hypothesis (28). Hence, current molecular genetic studies in humans do not definitively document a genetic component in the pathogenesis of LVH, whereas in experimental animals inherited susceptibility to LVH is well documented (29). Given the present data on human subjects, the overall familial contribution of LVH can be estimated and, thus, more realistic power estimations (to confirm or reject a null hypothesis of genetic association or linkage) for molecular genetic analyses can be made.
Conclusions. Although a familial and presumably a genetic component in the pathogenesis of LVH is suggested by this study, we do not wish to distract attention from measures that have been proven to regress LVH, including diet, physical activity, and, if indicated, antihypertensive medication. Furthermore, it remains to be shown that LVH derived from genetic predisposition confers an increased cardiovascular risk as has been documented for LVH in general (24,25). Nevertheless, the present data allow the reemphasis that several cardiovascular sequelae including LVH are potentially shared by family members of affected individuals, providing the opportunity for early recognition and, possibly, prevention of the associated risk.
| Acknowledgments |
|---|
| Footnotes |
|---|
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
H Schunkert and J Erdmann Well kept secrets of the genome Eur. Heart J., March 2, 2003; 24(6): 501 - 503. [Full Text] [PDF] |
||||
![]() |
B. M. Mayosi, B. Keavney, A. Kardos, C. H. Davies, P. J. Ratcliffe, M. Farrall, and H. Watkins Electrocardiographic measures of left ventricular hypertrophy show greater heritability than echocardiographic left ventricular mass Eur. Heart J., December 2, 2002; 23(24): 1963 - 1971. [Abstract] [PDF] |
||||
![]() |
K. Sedlacek, M. Fischer, J. Erdmann, C. Hengstenberg, S. Holmer, S. Kurzinger, M. Muscholl, A. Luchner, G. A. Riegger, H.-W. Hense, et al. Relation of the G Protein {beta}3-Subunit Polymorphism With Left Ventricle Structure and Function Hypertension, August 1, 2002; 40(2): 162 - 167. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. A. Oyama and W. P. Thomas Two-Dimensional and M-Mode Echocardiographic Predictors of Disease Severity in Dogs With Congenital Subaortic Stenosis J. Am. Anim. Hosp. Assoc., May 1, 2002; 38(3): 209 - 215. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. A. Kotchen, J. M. Kotchen, C. E. Grim, V. George, M. L. Kaldunski, A. W. Cowley, P. Hamet, and T. H. Chelius Genetic Determinants of Hypertension : Identification of Candidate Phenotypes Hypertension, July 1, 2000; 36(1): 7 - 13. [Abstract] [Full Text] [PDF] |
||||
![]() |
Y. Jamshidi, H. E. Montgomery, H.-W. Hense, S. G. Myerson, I. P. Torra, B. Staels, M. J. World, A. Doering, J. Erdmann, C. Hengstenberg, et al. Peroxisome Proliferator-Activated Receptor {alpha} Gene Regulates Left Ventricular Growth in Response to Exercise and Hypertension Circulation, February 26, 2002; 105(8): 950 - 955. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | SUBSCRIPTIONS | CURRENT ISSUE | PAST ISSUES | CARDIOSOURCE | SEARCH | HELP | FEEDBACK |