Correspondence

Absence of Association or Genetic Linkage between the Angiotensin-Converting–Enzyme Gene and Left Ventricular Mass

N Engl J Med 1996; 335:1070-1071October 3, 1996

Article

To the Editor:

We would like to comment on the study by Lindpaintner et al. (April 18 issue)1 reporting no association or genetic linkage between the angiotensin-converting–enzyme (ACE) gene and left ventricular mass.

Genetic linkage between the ACE locus and left ventricular mass was investigated with the use of a sibling-pair analysis. For quantitative traits, this method is known to have a very low power, unless the locus under consideration has a very large effect (heritability of 50 percent or more).2,3 The authors indicate that left ventricular mass has been demonstrated in the Framingham Heart Study to be heritable (their reference 48), but unfortunately they do not provide any estimate of this heritability. Yet this estimate would have been very useful in assessing the power of their study, provided that the heritability could be entirely attributable to the ACE locus. With the use of the power formulas given by Amos et al.3 and assuming a recombination fraction of 0.019 between the HGH microsatellite and the ACE gene, the power of a sibling-pair analysis of 759 pairs would be 40 percent to detect a genetic effect of 20 percent on the variance of left ventricular mass and 70 percent to detect a genetic effect of 30 percent, values that are substantially lower than those given by the authors. Our power calculations represent optimal values assuming complete knowledge of parental genotypes, independence of sibling pairs sampled from multiplex sibships, no dominance effect, and a susceptibility gene with a frequency of 0.5. Above all, a heritability of 30 percent of the left ventricular mass attributable to the ACE gene appears quite unrealistic, given the many factors that cause this phenotype and the likely underlying genetic heterogeneity. Moreover, major determinants of left ventricular mass, such as sex, age, body-mass index, and blood pressure, may influence the expression of genetic factors, and in the presence of such interaction even larger samples would be required.

The association study reported by the authors obviously had a much greater power than the linkage study to detect a possible effect of the ACE gene on left ventricular mass. Possible selection biases should be seriously considered when one is interpreting the results, however, knowing that 51 percent of the subjects who had echocardiography between 1979 and 1983 were not included in the present study. Left ventricular hypertrophy has been shown, in particular in the Framingham study,4 to be a strong predictor of cardiovascular mortality. It is striking that the prevalence of left ventricular hypertrophy was 18.7 percent in the 2740 women initially examined5 and only 14.9 percent in the 1323 women for whom ACE I-D polymorphism was genotyped, leaving a prevalence of 22.2 percent in the women not included in the present study (difference, P<0.001). Even though this difference was less marked in men (14.0 percent vs. 17.8 percent, P<0.02), it is clear that the exclusion of a fraction of subjects with a higher prevalence of left ventricular hypertrophy may have reduced the chance of detecting a true genetic effect. This selection may even have had a greater effect on the sibling-pair analysis, since the exclusion of siblings with extreme values considerably decreases the power of such an approach.2

François Cambien, M.D.
INSERM SC7, 75005 Paris, France

Laurence Tiret, Ph.D.
INSERM U258, 75674 Paris, France

5 References

1. 1

   Lindpaintner K, Lee M, Larson MG, et al. Absence of association or genetic linkage between the angiotensin-converting-enzyme gene and left ventricular mass. N Engl J Med 1996;334:1023-1028
   Full Text | Web of Science | Medline

2. 2

   Carey G, Williamson J. Linkage analysis of quantitative traits: increased power by using selected samples. Am J Hum Genet 1991;49:786-796
   Web of Science | Medline

3. 3

   Amos CI, Elston RC, Wilson AF, Bailey-Wilson JE. A more powerful robust sib-pair test of linkage for quantitative traits. Genet Epidemiol 1989;6:435-449
   CrossRef | Web of Science | Medline

4. 4

   Levy D, Garrison RJ, Savage DD, Kannel WB, Castelli WP. Prognostic implications of echocardiographically determined left ventricular mass in the Framingham Heart Study. N Engl J Med 1990;322:1561-1566
   Full Text | Web of Science | Medline

5. 5

   Levy D, Anderson KM, Savage DD, Balkus SA, Kannel WB, Castelli WP. Risk of ventricular arrhythmias in left ventricular hypertrophy: the Framingham Heart Study. Am J Cardiol 1987;60:560-565
   CrossRef | Web of Science | Medline

To the Editor:

Dr. Lindpaintner and his colleagues showed that the ACE gene is not a major determinant of left ventricular mass. Since we know that multiple factors, including obesity and blood pressure, affect left ventricular mass, the new finding is of some physiologic interest. However, the important clinical issue is whether there is any association between the ACE gene and left ventricular hypertrophy, because the latter is an extremely powerful risk factor for cardiovascular mortality. In this regard, although the study of Lindpaintner et al. has adequate power to detect small changes in the mean left ventricular mass of a population, it has insufficient power to detect small but important differences in the prevalence of left ventricular hypertrophy, since only 15.6 percent of their subjects had the condition. The wide confidence intervals for the odds ratios meant that the study could not rule out a 58 percent excess in left ventricular hypertrophy in subjects with the DD genotype as compared with those with the II genotype. Identifying subjects with extreme phenotypes is a powerful strategy in genetic investigations. It is noteworthy that the previous studies that reported a positive association between left ventricular hypertrophy and ACE gene polymorphism focused on subjects with left ventricular hypertrophy.1,2 Their use of electrocardiographic criteria for left ventricular hypertrophy enabled them to accurately define extreme phenotypes with high specificity (98.8 percent).3

Finally, if there is indeed no association or only a weak one between DD and left ventricular hypertrophy, then these two common risk factors in combination may increase cardiovascular risk markedly. The possible excessive mortality of people with these risk factors must surely be a matter of concern, especially since their premature deaths may lead to their underrepresentation in studies like the present one.

Bernard Cheung, M.B., B.Chir., Ph.D.
University of Hong Kong, Hong Kong

3 References

1. 1

   Schunkert H, Hense H-W, Holmer SR, et al. Association between a deletion polymorphism of the angiotensin-converting-enzyme gene and left ventricular hypertrophy. N Engl J Med 1994;330:1634-1638
   Full Text | Web of Science | Medline

2. 2

   Iwai N, Ohmichi N, Nakamura Y, Kinoshita M. DD genotype of the angiotensin-converting-enzyme gene is a risk factor for left ventricular hypertrophy. Circulation 1994;90:2622-2628
   Web of Science | Medline

3. 3

   Levy D, Labib SB, Anderson KM, Christiansen JC, Kannel WB, Castelli WP. Determinants of sensitivity and specificity of electrocardiographic criteria for left ventricular hypertrophy. Circulation 1990;81:815-820
   CrossRef | Web of Science | Medline

Author/Editor Response

The authors reply:

To the Editor: We agree with Drs. Cambien and Tiret that quantitative-trait analysis in sibling pairs has limited power, particularly for loci with modest heritabilities, and we have emphasized this in our paper. It is important to note that power estimates for genetic analyses are determined in part by the specified underlying genetic model and are appropriate only to the extent that the specified model reflects the unknown true state of nature. Given the previous reports of an association between the ACE D allele and left ventricular hypertrophy,1 we assumed a one-locus model, where the marker locus itself is responsible for variation in the trait. Cambien and Tiret correctly note that estimated power would be lower if one assumed a two-locus model using the method of Amos et al.2 With that method, our power estimates for a sibling-pair test between ACE and left ventricular mass using simulation (assuming 250 families with three siblings each, a recombination fraction of 0.0, equiprobable allele frequencies, and no linkage disequilibrium) are 0.26 and 0.43 for heritabilities of 20 percent and 30 percent, respectively. The number of informative matings, and thus the power, should be higher under a one-locus model, since every backcross mating at the trait locus is also, by definition, a backcross mating at the marker locus. Given the high degree of power provided by the concurrent association analyses, we believe that these considerations do not materially affect the overall interpretation of our study. The reasons for noninclusion of subjects in the study were noted in detail in the article, as was our particular concern over the 774 subjects who died or were lost to follow-up in the interval between the echocardiographic examinations and DNA collection. We also indicated that only by assuming that the effects of ACE on left ventricular mass or left ventricular hypertrophy had a major impact and were present exclusively in these subjects would the overall results of the study be altered — a possibility we considered unlikely. The absence of age-correlated shifts in ACE allele frequency among Framingham subjects may be viewed as further evidence against bias-by-attrition; this observed absence of excessive mortality of DD subjects also addresses one of Dr. Cheung's concerns. We appreciate his other comments concerning the power of comparing extreme phenotypes but note that one of the studies he cites1 was based, like ours, on a cross-sectional design with similar prevalence of left ventricular hypertrophy, and that the number of subjects with left ventricular hypertrophy analyzed in our study was 5 and 11 times higher, respectively, than in these two earlier studies,1,3 providing increased power.

Klaus Lindpaintner, M.D.
Brigham and Women's Hospital, Boston, MA 02115

Martin G. Larson, S.D.
Framingham Heart Study, Framingham, MA 01702

Alexander F. Wilson, Ph.D.
National Center for Human Genome Research, Baltimore, MD 21224

3 References

1. 1

   Schunkert H, Hense H-W, Holmer SR, et al. Association between a deletion polymorphism of the angiotensin-converting-enzyme gene and left ventricular hypertrophy. N Engl J Med 1994;330:1634-1638
   Full Text | Web of Science | Medline

2. 2

   Amos CI, Elston RC, Wilson AF, Bailey-Wilson JE. A more powerful robust sib-pair test of linkage for quantitative traits. Genet Epidemiol 1989;6:435-449
   CrossRef | Web of Science | Medline

3. 3

   Iwai N, Ohmichi N, Nakamura Y, Kinoshita M. DD genotype of the angiotensin-converting enzyme gene is a risk factor for left ventricular hypertrophy. Circulation 1994;90:2622-2628
   Web of Science | Medline