Donate Help Contact The AHA Sign In Home
American Heart Association
Circulation
Search: search_blue_button Advanced Search
« Previous Article | Table of Contents | Next Article »
Circulation. 1995;92:3240-3248

This Article
Right arrow Abstract Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Mahaney, M. C.
Right arrow Articles by MacCluer, J. W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Mahaney, M. C.
Right arrow Articles by MacCluer, J. W.

(Circulation. 1995;92:3240-3248.)
© 1995 American Heart Association, Inc.

Articles

Plasma HDL Cholesterol, Triglycerides, and Adiposity

A Quantitative Genetic Test of the Conjoint Trait Hypothesis in the San Antonio Family Heart Study

Michael C. Mahaney, PhD; John Blangero, PhD; Anthony G. Comuzzie, PhD; John L. VandeBerg, PhD; Michael P. Stern, MD; Jean W. MacCluer, PhD

From the Department of Genetics, Southwest Foundation for Biomedical Research (M.C.M., J.B., A.G.C., J.L.B, J.W.M.) and Division of Epidemiology, University of Texas Health Science Center at San Antonio (M.P.S.).


*    Abstract
up arrowTop
 *Abstract
down arrowIntroduction
 down arrowMethods
down arrowResults
down arrowDiscussion
down arrowReferences
 
Background The conjoint trait hypothesis proposes that combined low HDL cholesterol (HDL-C) and high triglyceride (TG) levels represent a single, inherited phenotype that adiposity may influence in an unspecified manner. We conducted formal statistical genetic tests of the conjoint trait hypothesis and the relation of the conjoint trait to adiposity using data for 569 subjects in 25 pedigrees from the San Antonio Family Heart Study.

Methods and Results We conducted multivariate genetic analyses to detect the effects of genes and environmental factors on variation in plasma concentrations of HDL-C and TG, fat mass (as percent body weight [FM%], determined by bioelectric impedance), and body mass index (BMI). We used maximum-likelihood methods to simultaneously estimate the phenotypic means and SDs, heritabilities (h2), effects of sex, age-by-sex, eight dietary and medical covariates, and genetic and environmental correlations. Likelihood ratio tests disclosed significant heritabilities (P<.001) for all traits (h2HDL-C=0.55, h2TG=0.53, h2FM%=0.37, h2BMI=0.44) but significant genetic correlations (P<.001), indicating pleiotropy, between two trait pairs only: HDL-C and TG ({rho}G=-0.52) and fat mass and BMI ({rho}G=0.86). We obtained significant environmental correlations between all trait pairs except HDL-C and BMI (P>.05).

Conclusions Both shared genes (pleiotropy) and shared environmental factors contribute to the commonly observed inverse phenotypic association between plasma levels of HDL-C and TG. Rather than low HDL-C and high TG being a single, genetically transmissible entity, it is the inverse relation between these two phenotypes throughout their normal ranges of variation as well as at the extremes that is influenced by shared genes and shared environments. However, common environmental factors, not shared genes, account for reported associations of plasma HDL-C and TG levels with measures of adiposity.


Key Words: genetics • lipids • lipoproteins • obesity


*    Introduction
up arrowTop
 up arrowAbstract
 *Introduction
down arrowMethods
down arrowResults
down arrowDiscussion
down arrowReferences
 
Plasma concentrations of HDL-C and TG generally are reported to be inversely related.1 2 Negative correlations ranging from -0.21 to -0.65 have been reported by the Framingham study,3 the Honolulu Heart Study,4 the Lipid Research Clinics Prevalence Study,5 and the Modesto, Calif study.6 Because there is a substantial loss of cholesteryl ester from the HDL particle as it becomes enriched with TG, elevated plasma TG levels have been suggested to be a major independent cause of depressed HDL-C levels.7 8 Although the exact mechanisms responsible for the relation are yet to be elucidated, the positive association of depressed plasma HDL-C and increased plasma TG with CVD is well known. Indeed, results from several clinical trials repeatedly have provided logical support for interventions that are directed toward altering multiple plasma lipid and/or lipoprotein levels, including elevating HDL-C and lowering TG, to reduce the risk of CVD.9 10 11 12

Measures of fatness and obesity also are associated with increased risk of CVD as well as non–insulin-dependent diabetes mellitus.13 14 15 16 Both metabolic and endocrinologic mechanisms undoubtedly contribute to the associations of obesity with various disease states and with the combination of low HDL-C and high TG.17 18 19 20

The association of low HDL-C with high TG is observed in many heritable disorders of lipoprotein metabolism,21 eg, familial combined dyslipidemia,22 23 familial hypertriglyceridemia,22 24 familial hypoalphalipoproteinemia,25 and familial dyslipoproteinemic hypertension.26 Similarly, increased adiposity has been observed in association with some of these disorders as well as with mutations that influence lipoprotein risk factors for CVD.27 28 29

Recently, in two studies using Cincinnati Lipid Research Clinic Family Study data, Sprecher et al30 31 advanced the hypothesis that low HDL-C and high TG occur conjointly and are transmitted across generations as a "combined phenotype" or "conjoint trait." They first arrived at this hypothesis after comparing the prevalence of combined HDL-C and TG abnormalities in first-degree relatives of probands with low HDL-C (<=10th percentile) only, high TG (>=90th percentile) only, and those with the low HDL-C/high TG trait. It was observed that first-degree relatives of low HDL-C/high TG probands were at increased risk of the combined phenotype compared with relatives of probands from the other two classes. In the subsequent study, Sprecher et al31 focused on a subsample from the same clinical population, which they enriched for hypertriglyceridemia by selecting probands with TG levels above the 95th percentile. Results obtained from a comparison of lipoprotein profiles for first-degree relatives of probands who were high TG/normal HDL-C with those of probands who were high TG/low HDL-C were interpreted to support the hypothesis that inheritance of a conjoint trait, rather than elevation of TG alone, was responsible for the well-known association. These studies also identified increased body mass, as measured by the Quetelet index (ie, QI=[weight/height2]x1000), as a factor that ". . . may be partially responsible for familial CT [conjoint trait]. . . ."30 While high TG/normal HDL-C and high TG/low HDL-C probands exhibited elevated QIs and those with the hypothesized conjoint trait were higher than those with high TG alone, the QIs of the first-degree relatives of these proband classes were not significantly different.

However, without a formal genetic analysis or test of the conjoint trait hypothesis, there is little basis for discrimination between the many possible causes for the conjoint, or more appropriately correlated, appearance of these traits in families. Many reasonable environmental and genetic scenarios can be invoked to account for familial aggregation of correlated low HDL-C, high TG, and obesity. As Sprecher et al31 recognize, the conclusion of a genetic basis for this correlation, even if warranted, still does not discriminate between the pleiotropic effects of additive polygenes and single loci or linkage. Given the range of reasonable functional scenarios for observed phenotypic associations between low HDL-C, high TG, and obesity, the shared effects of genes, or pleiotropy, is a much more parsimonious hypothesis for the conjoint inheritance of the two traits than is linkage. While the concept of pleiotropy has generally been applied to the multiple phenotypic effects of single loci, polygenic systems have also been shown to exhibit pleiotropic effects in quantitative characters.32 33

To discern whether quantitative variation in plasma levels of HDL-C and TG and measures of adiposity are inherited as a single conjoint trait, it is necessary to determine the proportion of the total phenotypic variance in HDL-C, TG, and measures of adiposity due to the additive effects of genes and the correlation between HDL-C, TG, and measures of adiposity due to shared genes and the effects of shared environment. We tested the conjoint trait hypothesis by conducting a multivariate statistical genetic analysis of quantitative variation in plasma HDL-C, plasma TG, and two measures reflective of obesity: BMI and fat mass.


*    Methods
up arrowTop
 up arrowAbstract
 up arrowIntroduction
 *Methods
down arrowResults
down arrowDiscussion
down arrowReferences
 
The data used in this study were obtained from a large sample of Mexican Americans, most of whom reside in San Antonio, Tex, who are participants in the San Antonio Family Heart Study, a broader project investigating the genetic determinants of atherosclerosis and its risk factors. The {approx}1250 participants in the San Antonio Family Heart Study include 40 probands, 40- to 60-year-old men and women residing in a low-income Mexican American barrio, and their first-, second-, and third-degree relatives who were >=16 years old. All probands were randomly ascertained with respect to disease status; the only eligibility criterion, apart from age, was that the proband had a spouse who also was willing to participate and six living first-degree relatives, excluding parents, who were >=16 years old. In addition to data on various lipoproteins, their subfractions, and enzymes and other proteins associated with them, extensive data pertaining to body composition, diet, nutrition, medications, physical activity, and socioeconomic status also have been collected. An Institutional Review Board (University of Texas Health Science Center at San Antonio) approved the data collection procedures, and all subjects gave informed consent.

For the present study, data were available for 569 individuals, 362 women and 207 men, whose ages ranged from 16 to 92 years (mean age, 39.4 years). Of these 569, 13 were unrelated to any other individual in the present sample, while the remaining individuals were distributed in 25 pedigrees. The sizes of these pedigrees ranged from 3 to 71 individuals, and all 25 pedigrees contained at least three generations of relatives. There are 117 sibships represented, ranging in size from 2 to 9, with a mean size of 3.02 and a modal size of 2 (n=56). This sample contains 3752 pairings of relatives for whom there are complete data on all variables. There are 850 first-degree relative pairs, including 464 sibling pairs, 261 mother-offspring pairs, and 125 father-offspring pairs; 1082 second-degree relative pairs, 1149 third-degree relative pairs, 570 fourth-degree relative pairs, and 92 fifth-degree relative pairs in the sample.

The study focused on four quantitative phenotypes: the two inversely related plasma phenotypes, HDL-C and TG, plus the two measures of relative weight and body composition, BMI and fat mass. Data on plasma concentrations of HDL-C (mmol/L) and TG (mmol/L) were obtained in accordance with standard clinical chemical methods34 from blood samples drawn in a clinical setting after a 12- to 14-hour fast. BMI was calculated as kg/m2 from weight (kg) and height (m) measurements, also obtained in the clinical setting by trained individuals following standard anthropometric protocols.35 BMI was included because of its common use in clinical and research settings as a measure of relative weight that is correlated with adiposity in adults. But, because the genetic and environmental correlations between these traits in this population were unknown before this study, we also included a more direct estimate of adiposity. FM% was estimated by means of bioelectrical impedance using a Valhalla 1990B body composition analyzer. The rationale, methodology, and utility of bioimpedance in body composition studies are well documented.36 37 38 39

In addition to age and sex, an extensive list of potential covariates of the four phenotypes was available for use in this study. These data were obtained from interview instruments used during the clinical visit that yielded the phenotype data. The instruments consisted of medical history interviews, which included personal and family history of diabetes, myocardial infarction, hypertension, stroke, vascular surgery, and current medications, and the Rose angina questionnaire40 ; and an environmental exposures interview, during which questionnaires on cigarette smoking, alcohol consumption, physical activity,41 and food frequency42 and dietary behavior43 were administered. In the analyses reported below, the additional covariates used included the ratio of dietary polyunsaturated fat to saturated fat and the percentage of dietary saturated fat (from the food frequency questionnaires) plus the following dichotomous (ie, 0,1) variables: postmenopausal status, exogenous sex hormones, diabetic status, diabetic medication use, cigarette smoking, and use of lipid-lowering medication.

Pedigree and phenotype data management and preparation were done with the computer package PEDSYS.44 Statistical genetic analyses used our modified version of the Pedigree Analysis Program, PAP, version 3.0,45 which uses maximum-likelihood methods to compute the likelihoods of genetic models on pedigrees.

According to classic quantitative genetic theory,46 the total phenotypic variance in a trait, {varsigma}2P, can be partitioned into {varsigma}2G, the variance due to the effects of genes, and {varsigma}2E, the variance due to environmental effects. These components are additive, such that


(1)

Each of these components can be further partitioned. Heritability, h2, the proportion of the total phenotypic variance due to the additive effects of genes, is obtained as {varsigma}2G/{varsigma}2P. On the basis of established quantitative genetic theory and method, it is possible to extend univariate genetic analysis to encompass the multivariate state.32 46 47 48 Following an approach presented in Williams-Blangero and Blangero33 and Comuzzie et al,49 we can model the multivariate phenotype of an individual as a linear function of the measurements on the individual's traits, the means of these traits in the population, the covariates and their regression coefficients, plus the additive genetic values and random environmental deviations. From such a model, we can obtain the phenotypic variance-covariance matrix, from which we can estimate the additive genetic and random environmental components, given the relationships (kinship coefficients) obtained from the pedigree and standard quantitative genetic theory. From the genetic and environmental variance-covariance matrices, it is a straightforward matter to estimate the additive genetic correlation, {rho}G, and the environmental correlation, {rho}E, between trait pairs.

Respectively, these two correlations are estimates of the effects of shared genes (ie, pleiotropy) and shared environmental factors on the phenotypic variance in a trait. Just as the genetic and environmental components of the phenotypic variance and covariance matrices for the pedigree are additive, so too are the components of the phenotypic correlation matrix. Therefore, by use of the maximum-likelihood estimates of additive genetic and environmental correlations, an estimate of the total phenotypic correlation between two traits, {rho}P, can then be obtained by means of the following identity:


(2)

Standard errors for the phenotypic correlations, obtained from Equation 2Up, were estimated by means of a first-order Taylor series approximation.

We conducted a quadrivariate quantitative genetic analysis of HDL-C, TG, BMI, and FM% using the simultaneous orthogonalization methods of Blangero and Konigsberg,50 implemented in our modified version of PAP version 3.0.45 This multivariate approach enabled simultaneous maximum-likelihood estimation of the 80 parameters, including the phenotypic means (µ), phenotypic standard deviations ({varsigma}), heritabilities (h2), and the effects of sex, age-by-sex, age2-by-sex, and the eight additional covariates (listed above) for all four traits, as well as the genetic and environmental correlations between them. Before analysis, HDL-C and TG were loge (ln) transformed to remove skewness in the data and mitigate the effects of scale on maximum-likelihood estimation. No other prior adjustments to the data were made.

While the quadrivariate model applied in this study is a polygenic one, it is possible that major genes could actually be involved in determining shared variation in the four phenotypes. The maximum-likelihood methods used in the study rely on the assumption of multivariate normality as a "working model" and are robust to deviations from multivariate normality in the underlying distribution. Consequently, valid maximum-likelihood estimates for the parameters of the genetic model can be obtained even if major loci, not modeled in this analysis, are involved.51

The significance of each of 68 of the estimated parameters (excluding µi and {varsigma}i) was assessed by likelihood ratio tests, in which -2xln likelihood of a restricted model, in which a parameter value is fixed at 0, is compared with the same statistic for the more general quadrivariate model in which all parameter values are estimated. The likelihood ratio test statistic, {Lambda}[i] (where i indicates degrees of freedom), is distributed approximately as a {chi}2 variate with degrees of freedom equal to the difference in the number of parameters in the two models being compared.52 Pleiotropy is indicated by additive genetic correlations that are found by likelihood ratio tests to be significantly different from zero. Genetic correlations were subjected to an additional likelihood ratio test, comparing the unrestricted model in which the correlation is estimated to restricted models, in which it is fixed at 1.0 or -1.0. This has been reported as a test of the extent of pleiotropy, with {rho}G significantly different from ||±1.0|| indicative of "incomplete" pleiotropy.53

Matrices of the maximum-likelihood estimates of the phenotypic, additive genetic, and random environmental correlations were each subjected to principal-components analysis to provide a graphical means of more readily appreciating the structures (ie, patterns of intercorrelations among the four phenotypes) of the three matrices. This approach allows us to represent the original four variables in a reduced multivariate space while maximizing the variance explained in the original data.54 This is accomplished by extracting the eigenvalues (ie, the characteristic values or latent roots) of a covariance matrix, the elements of which have been standardized to have means of 1 and variances of 0 (in other words, a correlation matrix). The eigenvalues indicate the variance explained by each of the principal components, which are the independent (ie, uncorrelated, orthogonal) linear functions of the original variates with coefficients given by the eigenvectors, nonzero scaling vectors corresponding to each eigenvalue. In principal-components analysis, the first eigenvalue explains the largest proportion of the variance in the data; the second, the next largest proportion; and so on. If the first two principal components account for a substantial proportion of the variance among the four phenoypes in this statistical genetic study, we find that plotting them against one another facilitates examination of their interrelations in two-dimensional space.


*    Results
up arrowTop
 up arrowAbstract
 up arrowIntroduction
 up arrowMethods
*Results
down arrowDiscussion
down arrowReferences
 
The univariate descriptive statistics for the raw measures for HDL-C, TG, BMI, and FM% are presented by sex in Table 1Down. Both female and male age- and sex-specific mean HDL-C and TG levels in the present study are within 1 SEM of their respective HHANES55 age- and sex-specific means, with 67% of the values (mean±SD) falling between the HHANES 15th and 85th percentiles for HDL-C and between the 15th and 90th percentiles for TG. Similar comparisons place the age- and sex-specific means from the present study intermediate to to HDL-C and TG values for US blacks and whites obtained by NHANES II investigators.56 Age- and sex-specific mean BMI values in the present study sample were at the 75th and 85th HHANES percentiles and at the 85th and 90th NHANES II percentiles for women and men, respectively.57 58


View this table:
[in this window]
[in a new window]
  		Table 1. Descriptive Statistical Summary for HDL-C, TG, BMI, and FM% in 569 Mexican Americans (362 Women, 207 Men) From the San Antonio Family Heart Study

Fig 1Down, in which the data for both sexes are combined, presents the univariate relative frequency distributions for the four phenotypes and the gaussian bivariate sample ellipses encompassing 86% of each pairwise distribution. The major axes of the ellipses are centered on the sample means, their lengths are determined by the unbiased SDs of the two phenotypes, and their orientations are functions of the sample covariances between them. Together, the univariate presentations in Table 1Up and Fig 1Down display evidence of the scale differences between HDL-C and TG and skewness of TG. Further, the gaussian bivariate sample ellipses in Fig 1Down provide initial graphic evidence of the phenotypic associations between the four traits. The shapes of the ellipses, reflecting the relative lengths of their major axes, and their orientations, reflecting the slopes of these axes, reveal an often observed negative association between HDL-C and TG and an expected positive association between BMI and FM%.



View larger version (29K):
[in this window]
[in a new window]
  		Figure 1. Scatterplot matrix of plasma lipoproteins and adiposity measures. Diagonal cells: Univariate relative frequency distributions of HDL-C, TG, BMI, and FM%. Off-diagonal cells: gaussian bivariate sample ellipses encompassing 86% (mean±1.5 SD) of the pairwise distributions.

The maximum-likelihood estimates of the mean effects and variance components and their SEEs from the quadrivariate quantitative genetic analysis of ln HDL-C, ln TG, BMI, and FM% are presented in Table 2Down. Likelihood ratio tests disclosed significant heritabilities for all four phenotypes, with point estimates of h2 for the plasma measures being about 25% higher than those of the two indicators of adiposity. Sex exerted significant effects on all phenotypes except BMI. Similarly, only in the case of ln HDL-C were the age terms nonsignificant contributors to the likelihood of the model. While the effect of the dichotomous postmenopausal status variable was not significant for any of the phenotypes, exogenous sex hormones did contribute to ln HDL-C and ln TG. Diabetic status but not the taking of diabetic medications had significant effects on ln TG, BMI, and FM%. The ratio of dietary polyunsaturated fat to saturated fat contributed significantly to the likelihood of the model but only in the case of ln TG. The effects of the remaining covariates (ie, smoking, percent of diet as saturated fat, lipid-lowering medication use) on the four phenotypes were not significant in this quantitative genetic model.


View this table:
[in this window]
[in a new window]
  		Table 2. Maximum Likelihood Parameter Estimates From a Quadrivariate Quantitative Genetic Analysis of ln HDL-C, ln TG, BMI, and FM% in 569 Mexican Americans From the San Antonio Family Heart Study

Table 3Down provides the maximum-likelihood estimates and SEEs for the remaining parameters of the quadrivariate quantitative genetic model: the additive genetic and environmental correlations between the four phenotypes, plus the phenotypic correlations, as estimated by Equation 2Up. Indications of pleiotropy, ie, significant additive genetic correlations, were observed only between ln HDL-C and ln TG, {rho}G=-0.523, and between BMI and FM%, {rho}G=0.864. No other genetic correlation contributed significantly to the likelihood of the model. Likelihood ratio tests also rejected the hypothesis of complete pleiotropy (ie, {rho}G=±1.00) for all genetic correlations (including that for BMI and FM%, where {chi}21=77.146, P<.000001). Nearly all the environmental correlations contributed significantly to the likelihood of the model. Only the environmental correlation between ln HDL-C and BMI, {rho}E=-0.187, was not significant ({chi}21=2.52, P=.1124). All environmental correlations between ln HDL-C and the other variables were negative, while all those between TG and the adiposity measures were positive.


View this table:
[in this window]
[in a new window]
  		Table 3. Maximum Likelihood Estimates of the Additive Genetic Correlations (G) and Environmental Correlations (E) From the Quadrivariate Quantitative Genetic Analysis of ln HDL-C, ln TG, BMI, and FM%, and the Total Phenotypic Correlations (P), Estimated as Indicated in Text

When squared, the significant additive genetic correlations estimate the proportion of the shared genetic variance attributable to the additive effects of genes. The squared additive genetic correlation between ln HDL-C and ln TG was {rho}G2=0.274±0.123 and between BMI and FM% was {rho}G2=0.747±0.085. The estimate of the proportion of the phenotypic variance accounted for by the phenotypic correlation between ln HDL-C and ln TG was {rho}P2=0.191±0.034 and between BMI and FM% was {rho}P2=0.576±0.030.

We conducted a principal-components analysis on the phenotypic, additive genetic, and random environmental correlation matrices to provide a graphical representation of the structure of each matrix. The first component axis commonly separates variables as a function of magnitude differences, while successive axes typically contrast variables on some other basis. The first and second principal component axes account for 98.1%, 99.5%, and 97.3% of the variation in the phenotypic, additive genetic, and random environmental correlation matrices, respectively. Graphical representations of the patterns of intercorrelations among the four phenotypes in each of the three matrices are presented in Fig 2aDown through 2c. In all three plots, ln HDL-C and ln TG were separated along both component axes, reflecting the large contributions of shared additive gene effects and shared random environments to the well-documented inverse phenotypic correlation between these two measures. The expected positive association between the two adiposity measures is evident in all three plots as well. The apparent phenotypic identity of these two traits (ie, the overlap in Fig 2aDown) results from the combined effects of shared additive genes (Fig 2bDown) and shared random environments (Fig 2cDown). However, the slight separation between the adiposity measures observed in Fig 2bDown and 2cDown confirms the results of likelihood ratio tests that rejected complete pleiotropy between FM% and BMI and suggests that a portion of the variation in each of these traits is affected by other unshared genes and other unshared random environmental effects. The two sets of significant pairwise correlations (ie, between ln HDL-C and ln TG and between FM% and BMI) are responsible for the similarities of pattern among the three plots. The differences in position, with respect to the adiposity measures, of ln HDL-C and ln TG in Fig 2bDown and 2cDown reflect sign changes in their nonsignificant correlations, rather than important differences between shared genetic and random environmental effects on the relations of the plasma lipid and lipoprotein measures to adiposity.



View larger version (9K):
[in this window]
[in a new window]
  		Figure 2. Plots of the first and second principal components obtained by decomposition of the phenotypic correlation matrix (a), the additive genetic correlation matrix (b), and the environmental correlation matrix (c).


*    Discussion
up arrowTop
 up arrowAbstract
 up arrowIntroduction
 up arrowMethods
up arrowResults
*Discussion
down arrowReferences
 
We used a quadrivariate quantitative genetic analysis to determine the contributions of genes, shared genes, and shared environments to normal human phenotypic variation in plasma HDL-C and TG, plus two measures of adiposity, BMI and FM%, in extended pedigrees constructed following the random ascertainment of probands. In the Mexican American population represented by this pedigreed sample from the San Antonio Family Heart Study, the commonly reported inverse phenotypic association between plasma concentrations of HDL-C and TG is due in large part both to the additive effects of shared genes and to shared environments. The same is true for the positive phenotypic association between BMI and FM%. In the case of HDL-C and TG, this observation of additive genetic pleiotropy is consistent with the observations by other researchers and clinicians of what has been referred to as a conjointly inherited trait. The significant genetic correlation between BMI and FM% confirms that, in this population, the majority of genes that influence phenotypic variation in BMI also influence phenotypic variation in fat mass. However, interpretation of the phenotypic associations between the two plasma phenotypes and measures of adiposity as evidence for the conjoint inheritance of low HDL-C, high TG, and obesity (whether measured as FM% or as wt/ht2) is not supported by the results of this study. Rather, those phenotypic associations are attributable to shared environmental effects and not to shared genes.

The present study is not the first to have implications for understanding the pleiotropic relations between lipid and/or lipoprotein measures or between this class of variables and indicators of adiposity. Research efforts by several groups have produced results indicative of pleiotropic interrelations among plasma-borne risk factors for cardiovascular disease. Path analytical methods, applied to assess the sources of variation and covariation in HDL, LDL, and VLDL in 160 nuclear families from the Cincinnati Lipid Research Clinic Family Study, rejected the hypothesis that common environmental effects alone could explain the phenotypic correlation between HDL and VLDL,59 suggested low to moderate genetic correlations between HDL and LDL and between LDL and VLDL, and estimated a strong environmental correlation between HDL and VLDL.60 With an analytical approach more similar to that of the present study, Blangero et al61 used multivariate maximum-likelihood segregation analysis to detect major locus pleiotropy and significant residual additive genetic and environmental correlations between a major locus for apolipoprotein A-I and five HDL-C subfractions in data from the Donner Laboratory Family Study. A quantitative genetic analysis of body mass, fat pattern, and serum lipid measures in 665 whites from 135 kindreds by Towne et al,53 also using maximum-likelihood methods, detected no significant additive genetic correlation between LDL-C and either BMI or waist-to-hip ratio. Towne et al53 did observe negative additive genetic correlations between HDL-C and BMI, which bordered on significance, and an additive genetic correlation between HDL-C and waist-to-hip ratio, which was significant. In a statistical genetic study of data from 2184 households conducted in Gubbio, Italy,62 variance decomposition analyses revealed that the phenotypic correlation between HDL-C concentration and BMI was attributable to shared random environmental and household components rather than to shared genetic effects.

The present study provides the first direct quantification of the genetic and environmental correlations between the components of the hypothesized combined trait, HDL-C and TG, as well as between HDL-C, TG, and measures of adiposity in extended pedigrees. It is the first formal test of the hypothesis of the conjoint inheritance of—or, more aptly, pleiotropy between—HDL-C and TG. Because the present study includes data from the normal range of variation for plasma HDL-C and TG levels, the detection of additive genetic correlations between these traits extends beyond support for the hypothesis that low HDL-C/high TG is a distinct transmissible entity. Our results show that it is the relation between HDL-C and TG, throughout their normal ranges of variation as well as at the extremes, that is heritable. Genes that are involved in the elevation of TG levels also are involved in lowering HDL-C and, conversely, genes that contribute to the elevation of HDL-C also contribute to the lowering of TG.

These conclusions are consistent with the reported metabolic interactions between HDL-C and TG. Many mechanisms that are known to lower plasma HDL-C concentrations are involved in TG metabolism,63 64 65 66 67 while mechanisms known to produce major alterations in TG metabolism also alter the size and concentration of HDLs present in the plasma.64 68 69 Heritable phenotypes hypothesized to mediate some of these interactions have been described.70

The shared genetic effects detected in the present study probably do not account for the relations between HDL-C and TG reported in all dyslipidemias. While more than 25% of the genetic variance in HDL-C and TG is attributable to shared genes, a portion of the remaining variation in plasma HDL-C and TG levels undoubtedly can be influenced independently by other unshared genes and/or environmental factors. Such additional influences could account for observations of uncorrelated responses between HDL-C and TG under specific physiological conditions or in association with specific deficiencies. For example, Yamashita et al71 reported elevated TG but normal HDL-C in individuals with reduced cholesteryl ester transfer protein activity, and Eisenberg and coworkers68 reported a similar association with alterations in the ratio of lipoprotein lipase to hepatic lipase.

The lack of genetic correlation between the lipid and adiposity measures should not be interpreted to mean that adiposity is unrelated to genes that may influence lipoprotein metabolism. The results of the present study indicate that, when the normal range of variation for plasma HDL-C, TG, and adiposity (whether measured as BMI or FM%) is included in the analysis, the phenotypic correlation of adiposity with HDL-C and TG is due to the effects of shared environmental factors rather than to the additive effects of genes shared by all of these traits. It is possible that HDL-C, TG, and adiposity are all genetically correlated to other traits not included in this study. A multivariate maximum-likelihood–based pedigree analysis that detected pleiotropy between a new phenotype and HDL-C, TG, and adiposity would not alter the patterns of intercorrelations already observed.

The heritability estimates obtained for these four traits in this pedigree sample are each very similar to the average of the heritabilities for each trait reported in twin, family, and pedigree studies of non-Hispanic groups from North America and Europe.72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 Detection of identical or nearly identical heritabilities for a trait in different populations must reflect similar relations between the additive genetic and environmental components of the total phenotypic variance in those populations. However, interstudy comparisons of the narrow sense heritability, h2, which is the estimate of the proportion of the total phenotypic variance due to the additive effects of genes, are of dubious utility. Many factors, including different ascertainment schemes, study designs, methods of parameter estimation, and population-specific environmental contributions to the phenotypic variance can result in dissimilar heritability estimates, even when the genetic variance estimates in the different populations are essentially the same. Identical heritability estimates for a trait in different populations are not necessary or sufficient to demonstrate involvement of identical genes in the expression of a trait, and dissimilar heritability estimates in different populations are not necessary or sufficient to exclude involvement of the same genes in the expression of the trait.

It is intergroup similarities in the phenotypes studied that provide useful insight regarding the generalizability of the results of any statistical genetic analysis from one population to another. Given that the majority of the measurements on the four phenotypes in the present study fall within normative standards for US populations of both similar and dissimilar ethnic composition and given the random ascertainment of the probands from which the San Antonio Familiy Heart Study pedigrees were reconstructed, we believe that these results, obtained from the statistical genetic analysis of data from 569 Mexican Americans predominantly residing in San Antonio, Tex, are generalizable to the US Mexican American population and beyond to other populations and ethnic groups.

In addition to providing insights into the patterns of genetic and environmental interactions among potentially related phenotypes, the multivariate approach used and the results obtained in this study have practical implications for other analyses. Large-magnitude genetic correlations obtained in multivariate quantitative genetic analyses can serve to delimit major locus hypotheses, including major locus pleiotropy and linkage, to be tested by multivariate segregation analysis and combined segregation and linkage analysis.93 Traits with high genetic or environmental correlations also can be used to great advantage as covariates in univariate segregation analysis. For example, on the basis of the results of the present study, we included TG as a covariate in a complex segregation analysis of plasma levels of HDL-C to account for shared additive genetic and random environmental contributions to the variance in HDL-C and TG levels. This helped to refine the HDL-C phenotype and increased the likelihood of the major gene model significantly.94

Plasma HDL-C and TG, the two components of the hypothesized conjoint trait, plus the measures of adiposity examined in the present study are all commonly accepted as complex phenotypes. To various extents, each of their observed patterns of quantitative variation is a function of interactions with genes and/or environmental factors that also influence quantitative variation in the other three phenotypes. Detecting pleiotropic and environmental interactions is an important first step in unraveling the determinants of phenotypic variation in cardiovascular risk factors such as HDL-C, TG, and adiposity. Statistical genetic models that in some way incorporate these interactions among phenotypes are more likely to approximate the biological reality of the phenotypes and consequently should be more likely to successfully detect, measure the effects of, localize, and identify genes contributing to commonly observed patterns of phenotypic variation and covariation.


*    Selected Abbreviations and Acronyms
 
BMI = body mass index
CVD = cardiovascular disease
FM% = fat mass expressed as a percentage of body weight
HDL-C = HDL cholesterol
TG = triglyceride(s)


*    Acknowledgments
 
This study was supported by National Institutes of Health (NIH) grant HL-45522. The development of statistical genetic methods used in this study was supported by NIH grants HL-28972, HL-45522, GM-31575, and DK-44297. The critical comments and insights of Dr D.L. Rainwater and Dr B.D. Mitchell and the programming assistance of T. Dyer are gratefully acknowledged.


*    Footnotes
 
Reprint requests to Michael C. Mahaney, PhD, Department of Genetics, Southwest Foundation for Biomedical Research, PO Box 28147, San Antonio, TX 78228-0147. E-mail mmahaney@darwin.sfbr.org.

Received April 24, 1995; revision received June 27, 1995; accepted July 24, 1995.


*    References
up arrowTop
 up arrowAbstract
 up arrowIntroduction
 up arrowMethods
up arrowResults
up arrowDiscussion
*References
 

  1. Brunzell JD, Austin MA. Plasma triglyceride levels and coronary heart disease. N Engl J Med. 1989;320:1273-1275. [Medline] [Order article via Infotrieve]
  2. Patsch W, Sharrett AR, Sorlie PD, Davis CE, Brown SA, ARIC Investigators. The relation of high density lipoprotein cholesterol and its subfractions to apolipoprotein A-I and fasting triglycerides: the role of environmental factors: the Atherosclerosis Risk in Communities (ARIC) study. Am J Epidemiol. 1992;136:546-557. [Abstract/Free Full Text]
  3. Gordon T, Castelli WP, Hjortland MC, Kannel WB, Dawber TR. High density lipoproteins as a protective factor against coronary heart disease. Am J Med. 1977;62:707-714. [Medline] [Order article via Infotrieve]
  4. Rhoads GG, Gulbrandsen CL, Kagan A. Serum lipoproteins and coronary heart disease in a population study of Hawaiian Japanese men. N Engl J Med. 1976;294:293-298. [Abstract]
  5. Davis CE, Gordon D, LaRosa J, Wood PDS, Halperin M. Correlations of plasma high density lipoprotein cholesterol levels with other plasma lipid and lipoprotein concentrations. Circulation. 1980;62(suppl IV):IV-24-IV-30.
  6. Albrink MJ, Krauss RM, Lindgren FT, Von Der Groeben VD, Wood PD. Intercorrelations among high density lipoprotein, obesity, and triglycerides in a normal population. Lipids. 1980;15:668-678. [Medline] [Order article via Infotrieve]
  7. Krauss RM. Regulation of high density lipoprotein levels. In: Havel RJ, ed. Medical Clinics of North America, Lipid Disorders. Philadelphia, Pa: WB Saunders Co; 1982;66:403-430.
  8. Nikkila EA, Taskinen MR, Sane T. Plasma high-density lipoprotein concentration and subfraction distribution in relation to plasma triglyceride metabolism. Am Heart J. 1987;113:543-550. [Medline] [Order article via Infotrieve]
  9. Carlson LA, Roshenhamer G. Reduction of mortality in the Stockholm Ischaemic Heart Disease Secondary Prevention Study by combined treatment with clofibrate and nicotinic acid. Acta Med Scand. 1988;223:405-418. [Medline] [Order article via Infotrieve]
  10. Manninen V, Elo MO, Frick MH, Haapa K, Heinonen OP, Heinsalmi P, Helo P, Huttunen JK, Kaitaniemi P, Koskinen P, Maenpaa H, Malkonen M, Mantarri M, Norola S, Pasternack A, Pikkarainen J, Romo M, Sjoblom T, Nikkila EA. Lipid alterations and decline in the incidence of coronary heart disease in the Helsinki Heart Study. JAMA. 1988;260:641-651. [Abstract]
  11. Blankenhorn DH, Nessim SA, Johnson RL, Sanmarco ME, Asen SP, Cashin-Hemphill L. Beneficial effects of combined colestipol-niacin therapy on coronary atherosclerosis and coronary venous bypass grafts. JAMA. 1987;257:3233-3240. [Abstract]
  12. Brown G, Albers JJ, Fisher LD, Schaefer SM, Lin JT, Kaplan C, Zhao XQ, Bisson BD, Fitzpatrick VF, Dodge HT. Regression of coronary artery disease as a result of intensive lipid lowering therapy in men with high levels of apolipoprotein B. N Engl J Med. 1990;323:1289-1298. [Abstract]
  13. Ohlson LO, Larsson B, Svardsudd K, Welin L, Eriksson H, Wilhelmsen L, Bjorntorp P, Tibblin G. The influence of body fat distribution on the incidence of diabetes mellitus: 13.5 years of follow-up of the participants in the study of men born in 1913. Diabetes. 1985;34:1055-1058. [Abstract]
  14. Stern MP, Haffner SM. Body fat distribution and hyperinsulinemia as risk factors for diabetes and cardiovascular disease. Arteriosclerosis. 1986;6:123-130. [Abstract/Free Full Text]
  15. Reichley KB, Mueller WH, Hanis CL, Joos SK, Tulloch BR, Barton S, Schull WJ. Centralized obesity and cardiovascular disease risk in Mexican Americans. Am J Epidemiol. 1987;125:373-386. [Abstract/Free Full Text]
  16. Larsson B, Bengtsson C, Bjorntorp P, Lapidus L, Sjoström L, Svardsudd K, Tibblin G, Wedel H, Welin L, Wilhelmsen L. Is abdominal body fat distribution a major explanation for sex difference in the incidence of myocardial infarction? Am J Epidemiol. 1992;135:266-273. [Abstract/Free Full Text]
  17. Haines AP, Imeson JD, Meade TW. Skinfold thickness and cardiovascular risk factors. Am J Epidemiol. 1987;126:86-94. [Abstract/Free Full Text]
  18. Lissner L, Sjöström L, Bengtsson C, Bouchard C, Larsson B. The natural history of obesity in an obese population and associations with metabolic aberrations. Int J Obes. 1994;18:441-447.
  19. Arai T, Yamashita S, Hirano K, Sakai N, Kotani K, Fujioka S, Nozaki S, Keno Y, Yamane M, Shinora E, Waliul Islam AHM, Ishigami M, Nakamura T, Kameda-Takemura K, Tokunaga K, Matsuzawa Y. Increased plasma cholesteryl ester transfer protein in obese subjects: a possible mechanism for the reduction of serum HDL cholesterol levels in obesity. Arterioscler Thromb. 1994;14:1129-1136. [Abstract/Free Full Text]
  20. Ward KD, Sparrow D, Vokonas PS, Willett WC, Landsberg L, Weiss ST. The relationships of abdominal obesity, hyperinsulinemia and saturated fat intake to serum lipid levels: the Normative Aging Study. Int J Obes. 1994;18:137-144.
  21. Tall AR. Plasma high density lipoproteins: metabolism and relationship to atherosclerosis. J Clin Invest. 1990;86:379-384.
  22. Goldstein JL, Schrott HG, Hazzard WR, Bieman EL, Motulsky AG. Hyperlipidemia in coronary heart disease, II: genetic analysis of lipid levels in 176 families and delineation of a new, inherited disorder, combined hyperlipidemia. J Clin Invest. 1973;52:1544-1568.
  23. Brunzell JD, Albers JJ, Chait A, Grundy SM, Groszek E, McDonald GB. Plasma lipoproteins in familial combined hyperlipidemia and monogenic familial hypertriglyceridemia. J Lipid Res. 1983;24:147-155. [Abstract]
  24. Brunzell JD, Schrott HG, Motulsky AG, Bierman EL. Myocardial infarction in the familial forms of hypertriglyceridemia. Metabolism. 1976;25:313-320. [Medline] [Order article via Infotrieve]
  25. Third JLHC, Montag L, Flynn M, Freidel J, Laskarzewski PM, Glueck CJ. Primary and familial hypoalphalipoproteinemia. Metabolism. 1984;33:136-146. [Medline] [Order article via Infotrieve]
  26. Williams RR, Hopkins PN, Hunt SC, Wu LL, Hasstedt SJ, Lalouel JM, Ash KO, Stults BM, Kuida H. Population-based frequency of dyslipidemia syndromes in coronary-prone families in Utah. Arch Intern Med. 1990;150:582-588. [Abstract]
  27. Morris BJ, Zee RYL, Robinson BG. Significant relationships of plasma lipids and body mass index with polymorphisms at the linked low-density-lipoprotein receptor gene and insulin receptor gene loci (19p13.2) in essential hypertensive patients. Clin Sci. 1994;86:583-592. [Medline] [Order article via Infotrieve]
  28. Vijayaraghavan S, Hitman GA, Kopelman PG. Apolipoprotein-D polymorphism: a genetic marker for obesity and hyperinsulinemia. J Clin Endocrinol Metab. 1994;79:568-570. [Abstract]
  29. Durack-Brown I, Fumeron F, Betouille D, Chevereuil O, Apfelbaum M. Xmn1 restriction polymorphism of apolipoprotein AI gene and lipoprotein levels in obesity. Int J Obes. 1994;18:460-464.
  30. Sprecher DL, Feigelson HS, Laskarzewski PM. The low HDL cholesterol/high triglyceride trait. Arterioscler Thromb. 1993;13:495-504. [Abstract/Free Full Text]
  31. Sprecher DL, Hein MJ, Laskarzewski PM. Conjoint high triglycerides and low HDL cholesterol across generations: analysis of proband hypertriglyceridemia and lipid/lipoprotein disorders in first-degree family members. Circulation. 1994;90:1177-1184. [Abstract/Free Full Text]
  32. Boehnke M, Moll PP, Kottke BA, Weidman WH. Partitioning the variability of fasting glucose levels in pedigrees. Am J Epidemiol. 1987;125:679-689. [Abstract/Free Full Text]
  33. Williams-Blangero S, Blangero J. Quantitative genetic analysis of skin reflectance: a multivariate approach. Hum Biol. 1992;64:35-49. [Medline] [Order article via Infotrieve]
  34. Lipid Research Clinics Program Manual of Laboratory Operations: Lipid and Lipoprotein Analysis, vol 1. Bethesda, Md: National Institutes of Health, 1974 (DHEW publication [NIH] 72-628).
  35. Lohman TG. Anthropometry and body composition. In: Lohman TG, Roche AF, Martorell R, eds. Anthropometric Standardization Reference Manual. Champaign, Ill: Human Kinetics Books; 1988:125-129.
  36. Nyboer J. Percent body fat by four terminal bio-electrical impedance and body density in college freshmen. In: Proceedings of the Vth International Conference on Electrical Bio-Impedance. Tokyo, Japan: August, 1981.
  37. Guo S, Roche AF, Chumlea WC, Miles DS, Pohlman R. Body composition predictions from bioelectric impedance. Hum Biol. 1987;59:221-233. [Medline] [Order article via Infotrieve]
  38. Baumgartner RN, Chumlea WC, Roche AF. Bioelectric impedance for body composition. Exerc Sports Sci Rev. 1990;18:93-224.
  39. Shephard RJ. Body Composition in Biological Anthropology. Cambridge Studies in Biological Anthropology. Cambridge, UK: Cambridge University Press; 1991:49-51.
  40. Rose GA, Blackburn H, Gillum RF, Prineas RJ. Cardiovascular Survey Methods. 2nd ed. Geneva, Switzerland: World Health Organization; 1982:162-165.
  41. Sallis JF, Haskell WL, Wood PD, Fortmann SP, Rogers T, Blair SN, Paffenbarger RS. Physical activity assessment methodology in the five-city project. Am J Epidemiol. 1985;121:91-106. [Abstract/Free Full Text]
  42. Willett W, Sampson L, Stampfer M, Rosner B, Bain C, Witschi S, Hennekens C, Speizer F. Reproducibility and validity of a semiquantitative food frequency questionnaire. Am J Epidemiol. 1985;122:51-65. [Abstract/Free Full Text]
  43. Knapp JA, Hazuda HP, Haffner SM, Young EA, Stern MP. A saturated fat/cholesterol avoidance scale: sex and ethnic differences in a biethnic population. J Am Dietet Assoc. 1988;88:172-177. [Medline] [Order article via Infotrieve]
  44. Dyke B. PEDSYS. A Pedigree Data Management System. San Antonio, Tex: Population Genetics Laboratory, Southwest Foundation for Biomedical Research; 1992.
  45. Hasstedt SJ. Pedigree Analysis Package, v 3.0. Salt Lake City, Utah: Department of Genetics, University of Utah; 1989.
  46. Falconer DS. Introduction to Quantitative Genetics. 3rd ed. Essex, UK: Longman; 1989.
  47. Hopper JL, Mathews JD. Extensions to multivariate normal models for pedigree analysis. Ann Hum Genet. 1982;46:373-383.
  48. Lange K, Boehnke M. Extensions to pedigree analysis, IV: covariance components models for multivariate traits. Am J Med Genet. 1983;14:513-524. [Medline] [Order article via Infotrieve]
  49. Comuzzie AG, Blangero J, Mahaney MC, Mitchell BD, Stern MP, MacCluer JW. Genetic and environmental correlations among skinfold measures. Int J Obes. 1994;18:413-418. [Medline] [Order article via Infotrieve]
  50. Blangero J, Konigsberg LW. Multivariate segregation analysis using the mixed model. Genet Epidemiol. 1991;8:299-316. [Medline] [Order article via Infotrieve]
  51. Beaty TH, Self SG, Liang KY, Connolly MA, Chase GA, Kwiterovich PO. Use of robust variance components models to analyse triglyceride data in families. Ann Hum Genet. 1985;49:315-328. [Medline] [Order article via Infotrieve]
  52. Edwards AWF. Likelihood. Cambridge, UK: Cambridge University Press; 1972.
  53. Towne B, Roche AF, Chumlea WC, Guo S, Siervogel RM. Analysis of pleiotropy between body mass, fat pattern, and serum HDL-C and LDL-C concentrations. Am J Phys Anthropol. 1994;suppl 18:196-197. Abstract.
  54. Manly BFJ. Multivariate Statistical Methods: A Primer. London, UK: Chapman & Hall; 1986:59-71.
  55. Carroll M, Sempos C, Fulwood R, Bachorik PS, Rifkind B, Lippel K. Serum Lipids and Lipoproteins of Hispanics, 1982-84. Hyattsville, Md: US Dept of Health and Human Services, Public Health Service, National Center for Health Statistics, 1987 (Vital and Health Statistics, Series 11: Data from the National Health Examination Survey and the National Health and Nutrition Examination Survey, No. 240. DHHS Publication [PHS] 90-1690).
  56. Carroll M, Sempos C, Briefel R, Gray S, Johnson C. Serum Lipids of Adults 20-74 Years: United States, 1976-1980. Hyattsville, Md: US Dept of Health and Human Services, Public Health Service, National Center for Health Statistics, 1990 (Vital and Health Statistics, Series 11: Data from the National Health Survey, No. 242. DHHS Publication [PHS] 93-1692).
  57. Najjar MF, Kuczmarski RJ. Anthropometric Data and Prevalence of Overweight for Hispanics: 1982-1984. Hyattsville, Md: US Dept of Health and Human Services, Public Health Service, National Center for Health Statistics, 1989 (Vital and Health Statistics, Series 11: Data from the National Health Survey, No. 239. DHHS Publication [PHS] 89-1689).
  58. Najjar MF, Rowland M. Anthropometric Data and Prevalence of Overweight, United States, 1976-1980. Hyattsville, Md: US Dept of Health and Human Services, Public Health Service, National Center for Health Statistics, 1987 (Vital and Health Statistics, Series 11: Data from the National Health Survey, No. 238. DHHS Publication [PHS] 87-1688).
  59. McGue M, Rao DC, Iselius I, Russell JM. Resolution of genetic and cultural inheritance in twin families by path analysis: application to HDL-cholesterol. Am J Hum Genet. 1985;37:998-1014. [Medline] [Order article via Infotrieve]
  60. Vogler GP, Rao DC, Laskarzewski PM, Glueck CJ, Russell JM. Multivariate analysis of lipoprotein cholesterol fractions. Am J Epidemiol. 1987;125:706-719. [Abstract/Free Full Text]
  61. Blangero J, Williams-Blangero S, Mahaney MC. Multivariate genetic analysis of Apo AI concentration and HDL subfractions: evidence for major locus pleiotropy. Genet Epidemiol. 1993;10:617-622. [Medline] [Order article via Infotrieve]
  62. Schork NJ, Weder AB, Trevisan M, Laurenzi M. The contribution of pleiotropy to blood pressure and body-mass index variation: the Gubbio study. Am J Hum Genet. 1994;54:361-373. [Medline] [Order article via Infotrieve]
  63. Eisenberg S. High density lipoprotein metabolism. J Lipid Res. 1984;25:1017-1018. [Medline] [Order article via Infotrieve]
  64. Eisenberg S, Brinton EA, Breslow JL. Triglyceride and high density lipoprotein. In: Carlson LA, ed. Disorders of HDL. Gunter Grove, London, UK: Smith Gordon & Co, Ltd; 1990;16:127-132.
  65. Tverdal A, Foss OP, Leren P, Holme I, Larsen-Lund PG, Bjartveit K. Serum triglycerides as an independent risk factor for death from coronary heart disease in middle-aged Norwegian men. Am J Epidemiol. 1989;129:458-465. [Abstract/Free Full Text]
  66. Brinton EA, Eisenberg S, Breslow JL. Human HDL cholesterol levels are determined by ApoA-I fractional catabolic rate, which correlates inversely with estimates of HDL particle size: effects of gender, hepatic and lipoprotein lipases, triglyceride and insulin levels, and body fat distribution. Arterioscler Thromb. 1994;14:707-720. [Abstract/Free Full Text]
  67. Pocock SJ, Shaper AG, Phillips AN. Concentrations of high density lipoprotein cholesterol, triglyceride and total cholesterol in ischaemic heart disease. Br Med J. 1989;298:998-1002.
  68. Blum CB, Levy RI, Eisenberg S, Hall M, Goebel RH, Berman M. High density lipoprotein metabolism in man. J Clin Invest. 1977;60:795-807.
  69. Frohlich JJ, Pritchard PH, Sparks DL, Westerlund J. Primary isolated hypo-alpha-lipoproteinemia: a distinct familial lipoprotein abnormality? In Carlson LA, ed. Disorders of HDL. Gunter Grove, London, UK: Smith Gordon & Co, Ltd; 1990;16:177-183.
  70. Sane T, Nikkila EA. Very low density lipoprotein triglyceride metabolism in relatives of hypertriglyceridemic probands: evidence for genetic control of triglyceride removal. Arteriosclerosis. 1988;8:217-226. [Abstract/Free Full Text]
  71. Yamashita S, Hui DY, Wetterau JR, Sprecher DL, Harmony JAK, Sakai N, Matsuzawa Y, Tarui S. Characterization of plasma lipoproteins in patients heterozygous for human plasma cholesteryl transfer protein (CETP) deficiency: plasma CETP regulates high density lipoprotein concentration and composition. Metabolism. 1991;40:756-763. [Medline] [Order article via Infotrieve]
  72. Duccini Dal Colletto GM, Krieger H, Magalhaes JR. Estimates of the genetical and environmental determinants of serum lipid and lipoprotein concentrations in Brazilian twins. Hum Hered. 1981;31:232-237. [Medline] [Order article via Infotrieve]
  73. Rao DC, Laskarewski PM, Morrison JA, Khoury P, Kelly K, Wette R, Russell J, Glueck CJ. The Cincinnati Lipid Research Clinic Family Study: cultural and biological determinants of lipids and lipoprotein concentrations. Am J Hum Genet. 1982;34:888-903. [Medline] [Order article via Infotrieve]
  74. Dahlen G, Ericson C, de Faire U, Iselius L, Lundman T. Genetic and environmental determinants of cholesterol and HDL-cholesterol concentrations in blood. Int J Epidemiol. 1983;12:32-35. [Abstract/Free Full Text]
  75. Whitfield JB, Martin NG. Plasma lipids in twins: environmental and genetic influences. Atherosclerosis. 1983;48:265-277. [Medline] [Order article via Infotrieve]
  76. Iselius L, Carlson LA, Morton NE, Efendic S, Lindsten J, Luft R. Genetic and environmental determinants for lipoprotein concentrations in blood. Acta Med Scand. 1985;217:161-170. [Medline] [Order article via Infotrieve]
  77. Namboodiri KK, Kaplan EB, Heuch I, Elston RC, Green PP, Rao DC, Laskarzewski P, Glueck CJ, Rifkind BM. The Collaborative Lipid Research Clinics Family Study: biological and cultural determinants of familial resemblance for plasma lipids and lipoproteins. Genet Epidemiol. 1985;2:227-254. [Medline] [Order article via Infotrieve]
  78. Hamsten A, Iselius L, Dahlen G, de Faire U. Genetic and cultural inheritance of serum lipids, low and high density lipoprotein cholesterol, and serum apolipoproteins A-I, A-II, and B. Atherosclerosis. 1986;60:199-208. [Medline] [Order article via Infotrieve]
  79. Austin MA, King MC, Bawol RD, Hulley SB, Friedman GD. Risk factors for coronary heart disease in adult female twins: genetic heritability and shared environmental influences. Am J Epidemiol. 1987;125:308-318. [Abstract/Free Full Text]
  80. Bucher KD, Friedlander Y, Kaplan EB, Namboodiri KK, Kark JD, Eisenberg S, Stein Y, Rifkind BM. Biological and cultural inheritance of familial resemblance in plasma lipids: a comparison between North America and Israel: the Lipid Research Clinics Program. Genet Epidemiol. 1988;5:17-33. [Medline] [Order article via Infotrieve]
  81. O'Connell DL, Heller RF, Roberts DCK, Allen JR, Knapp JC, Steele PL, Silove D. Twin study of genetic and environmental effects on lipid levels. Genet Epidemiol. 1988;5:323-341. [Medline] [Order article via Infotrieve]
  82. Hunt SC, Hasstedt SJ, Kuida H, Stults BM, Hopkins PN, Williams RR. Genetic heritability and common environmental components of resting and stressed blood pressures, lipids and body mass index in Utah pedigrees and twins. Am J Epidemiol. 1989;129:625-638. [Abstract/Free Full Text]
  83. Perusse L, Despres JP, Tremblay A, Leblanc C, Talbot J, Allard C, Bouchard C. Genetic and environmental determinants of serum lipids and lipoproteins in French Canadian families. Arteriosclerosis. 1989;9:308-318. [Abstract/Free Full Text]
  84. Rice T, Vogler GP, Perry TS, Laskarzewski PM, Rao DC. Familial aggregation of lipids and lipoproteins in families ascertained through random and nonrandom probands in the Iowa Lipid Research Clinics Family Study. Hum Hered. 1991;41:107-121. [Medline] [Order article via Infotrieve]
  85. Cullen P, Farren B, Scott J, Farrall M. Complex segregation analysis provides evidence for a major gene acting on serum triglyceride levels in 55 British families with familial combined hypertriglyceridemia. Arterioscler Thromb. 1994;14:1233-1249. [Abstract/Free Full Text]
  86. Burns TL, Moll PP, Lauer RM. The relation between ponderosity and coronary risk factors in children and their relatives: the Muscatine Ponderosity Family Study. Am J Epidemiol. 1989;129:973-987. [Abstract/Free Full Text]
  87. Stunkard AJ, Harris JR, Pedersen NL, McCearn GE. The body-mass index of twins who have been reared apart. N Engl J Med. 1990;322:1483-1487. [Abstract]
  88. MacDonald A, Stunkard AJ. Body-mass indexes of British separated twins. N Engl J Med. 1990;322:1530. Letter. [Medline] [Order article via Infotrieve]
  89. Moll PP, Burns TL, Lauer RM. The genetic and environmental sources of body mass index variability: the Muscatine Ponderosity Family Study. Am J Hum Genet. 1991;49:1243-1255. [Medline] [Order article via Infotrieve]
  90. Bouchard C, Pérusse L, LeBlanc C, Tremblay A, Thériault G. Inheritance of the amount and distribution of body fat. Int J Obes. 1988;12:205-215. [Medline] [Order article via Infotrieve]
  91. Price RA, Gottesman I. Body fat in identical twins reared apart: roles for genes and environment. Behav Genet. 1991;21:1-7. [Medline] [Order article via Infotrieve]
  92. Rice T, Borecki IB, Bouchard C, Rao DC. Segregation analysis of fat mass and other body composition measures derived from underwater weighing. Am J Hum Genet. 1993;52:967-973. [Medline] [Order article via Infotrieve]
  93. Mahaney MC, Jaquish CE, Comuzzie AG. Statistical genetics of normal variation in family data for oligogenic diseases. Genet Epidemiol. In press.
  94. Mahaney MC, Blangero J, Comuzzie AG, Rainwater DL, VandeBerg JL, Stern MP, MacCluer JW. Evidence for a major gene influencing normal quantitative variation in high density lipoprotein cholesterol (HDL-C) in Mexican Americans. Am J Hum Biol. 1995;7:129-130. Abstract.



This article has been cited by other articles: