© 2000 American Heart Association, Inc.
Atherosclerosis and Lipoproteins |
Segregation Analysis of Apolipoproteins A-1 and B-100 Measured Before and After an Exercise Training Program
The HERITAGE Family Study
From the Division of Biostatistics (P.A., T.R., I.B.B., D.C.R.) and Department of Genetics and Psychiatry (D.C.R.), Washington University School of Medicine, St Louis, Mo; Laboratory of Molecular Endocrinology, CHUL Research Center (J.G.) and Lipid Research Center (J.B., J.-P.D.), Laval University, Québec, Canada; School of Kinesiology and Leisure Studies, University of Minnesota, Minneapolis (A.S.L.); Department of Kinesiology, Indiana University, Bloomington (J.S.S.); Department of Health and Kinesiology, Texas A&M University, College Station (J.H.W.); and Pennington Biomedical Research Center, Louisiana State University, Baton Rouge (C.B.).
Correspondence to Ping An, MD, Division of Biostatistics, Campus Box 8067, Washington University School of Medicine, 660 S Euclid Ave, St Louis, Mo 63110-1093. E-mail anping{at}wubios.wustl.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Abstract—Complex segregation analyses of apolipoproteins (apo) A-1 and B-100 were performed in a sample of 520 individuals from 99 white families who participated in the HERITAGE Family Study. In these sedentary families, plasma apo A-1 and B-100 concentrations were measured before and after a 20-week endurance exercise training program. Baseline apo A-1 and B-100 were adjusted for the effects of age (age-adjusted baseline apo A-1 and B-100) and for the effects of age and BMI (age-BMI–adjusted baseline apo A-1 and B-100). The change in response to training was computed as a simple
(posttraining minus baseline) and was adjusted for age and
the baseline (age-baseline–adjusted apo A-1 and B-100 responses to
training). In the present study, a major gene could not be inferred
for baseline apo A-1. Rather, we found a major effect along with a
multifactorial effect accounting for 8% to 9% and 51% to 56% of the
variance, respectively. In addition, no clear evidence supported a
major-gene effect for its response to training, whereas the
transmission of a major effect from parents to offspring was ambiguous,
ie, genetic in nature or familial environmental in origin. The major
effect accounted for 15% of the variance, with an additional 21% and
58% of the variance being accounted for by a multifactorial effect in
parents and offspring, respectively. It is interesting to have obtained
evidence of a putative recessive major locus for baseline apo B-100,
which accounted for 50% to 56% of the variance, with an additional
25% to 29% of the variance due to a multifactorial effect. In
contrast, no major effect for its response to training was identified,
although a multifactorial effect was found that accounted for 27% of
the variance. The novel findings arising from the present study are
summarized as follows. Baseline apo A-1 and its response to training
were influenced by a major effect and a multifactorial effect. Baseline
apo B-100 was influenced by a putative major recessive gene with a
multifactorial component, but its response to training was influenced
solely by a multifactorial component in these sedentary families.
Key Words: commingling • major gene effect • major effect • multifactorial effect
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Some evidence indicates that elevated plasma and total LDL cholesterol (LDL-C) levels are major etiologic factors for developing atherosclerosis and coronary heart disease, whereas elevated HDL cholesterol (HDL-C) concentrations appear to provide protective effects against premature coronary heart disease. Apolipoprotein (apo) A-1 is one of the major protein components of HDL. It activates lecithin:cholesterol acyltransferase (LCAT), a plasma enzyme that catalyzes the esterification of cholesterol on the HDL particle surface, making possible reverse cholesterol transport from peripheral tissues.1 Apo B-100 is predominant in LDL and VLDL, which transport cholesterol and triglycerides to peripheral tissues in the fasting state. Apo B-100, the sole lipoprotein in LDL, serves as the ligand for binding to cell LDL receptors in the liver and other tissues.2
The clinical significance of apo A-1 and B-100 has led to extensive studies for 2 decades focusing on the level and nature of genetic influences.3 4 The familial aggregation of apo A-1 is well documented, with genetic heritability estimates reported to range from 53% to 66% in twin studies5 6 and to reach 43% in families selected through probands with premature myocardial infarction and in families randomly selected from the general population.7 For apo B-100, Hamsten et al7 used path analysis in families and reported genetic heritabilities of 14% in parents and 51% in offspring. Results obtained from segregation analyses have suggested major-gene effects for both apo A-18 9 10 11 12 13 14 15 16 and apo B-100.2 12 17 18 19 20 More recent studies have reported a second major locus determining apo A-1 levels.21 22 In contrast to these positive findings, there also have been contradictory reports for apo A-19 18 23 and apo B-100.24 25
Prospective epidemiological studies have provided evidence that apolipoprotein levels can be affected by physical activity and fitness levels.26 In particular, apo A-1 levels have been reported to increase and apo B-100 levels to decrease27 28 in response to exercise training. However, no studies indicated whether there is a familial basis for the changes in response to exercise training.
Physical activity level was controlled for in the HEalth, RIsk factors, exercise Training and GEnetics (HERITAGE) Family Study by requiring all participants to be sedentary at baseline, ie, having not engaged in regular physical activity over the previous 6 months. Apo A-1 and B-100 were measured before and after a 20-week endurance exercise training program. The present study is unique in that the contribution of physical activity level to the baseline apo A-1 and B-100 variability is somewhat reduced, because the HERITAGE Family Study participants were required to be sedentary at entry. Furthermore, this study is unique in that apo A-1 and B-100 levels were assessed before and after exercise training in intact families, so that genetic hypotheses regarding the response to exercise training can be evaluated.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Sample
The HERITAGE Family Study was designed to investigate the role of the genotype in cardiovascular, metabolic, and hormonal responses to aerobic exercise training and the contribution of exercise training to changes in several cardiovascular disease and diabetes risk factors. See Bouchard et al29 for more details concerning the HERITAGE Family Study sample and protocol.
A total of 520 individuals from 99 white families (255 men, 265
women) were analyzed in this study. Participants with
incomplete baseline or postexercise training apo A-1 and B-100 measures
were excluded from these analyses. Table 1
gives the sample sizes within
each of 4 sex-by-generation subgroups (fathers, mothers, sons, and
daughters) for the baseline measurements and the changes in response to
exercise training. Black families were also recruited for the HERITAGE
Family Study but were not included in these analyses because of
inadequate sample size. Recruitment of families was based on extensive
media publicity and advertisements at the 4 participating clinical
centers.
|
The following criteria were applied to screen subjects for
participation. First, individuals had to be between the ages of 17 and
65 years (17 and 40 years for children and 
65 years for parents).
Second, all participants were required to be sedentary at baseline.
Third, individuals with a body mass index [BMI, weight (kg)/height
(m)2] >40 kg/m2 were
excluded unless it was determined during baseline testing that they
were able to meet the demands of the exercise tests and exercise
training program. Fourth, resting blood pressure (BP) levels could not
exceed 159 mm Hg for systolic BP and 99 mm Hg for
diastolic BP without medications. Finally, participants
were required to be in good general physical health to complete the
20-week exercise training program. Further exclusion criteria can be
found in Bouchard et al.29
Exercise Training Program
The training protocol is thoroughly outlined in Bouchard et
al.29 Briefly, each individual trained on a cycle
ergometer in the laboratory under supervision 3 times a week for 20
weeks. The intensity and duration of exercise were adjusted for each
individual every 2 weeks so that each participant was working at a
heart rate associated with 75% of their maximum oxygen uptake volume
for 50 minutes at the last 6 weeks of training. The power output was
adjusted automatically via built-in computers in the cycle ergometer to
meet the designed heart rate response to exercise training. All
training sessions were supervised on site, and adherence to the
protocol was strictly monitored.
Measurements
Before and after the 20-week standardized exercise training
program, blood samples were taken from an antecubital vein into vacuum
tubes containing EDTA to measure lipid levels, including plasma apo A-1
and B-100 concentrations. The blood samples were collected in the
morning in a 12-hour fasting state with the participants in a
semirecumbent position and were drawn twice before (24 hours apart) and
twice after (24 and 72 hours after the last exercise session) a 20-week
endurance exercise training program. The blood samples collected at
each clinical center were prepared according to a standard protocol
before being sent to the core laboratory in Québec and were
adjusted for possible hemodilution by protein assays (total proteins
were assayed by the Biuret method (Roche Molecular Biochemicals,
Boehringer Mannheim Corp). Apo A-1 and B-100 concentrations
were tested in plasma by the rocket immunoelectrophoretic method of
Laurell.30 See Després et al31 for more
details on the internal and external quality control of plasma apo A-1
and B-100 assays.
Data Adjustments
Baseline apo A-1 and B-100 were adjusted for the effects of age,
separately within each of the 4 sex-by-generation groups, by a stepwise
multiple-regression procedure. Briefly, a given measurement was
regressed on a polynomial in age (linear, quadratic, and cubic) in a
stepwise manner, retaining only those terms that were significant at
the 5% level. Thus, the residual score from this regression should be
independent of age, sex, and generation effects. Similar sets of
stepwise regressions (by sex and generation groups) were performed on
baseline apo A-1 and B-100 with a polynomial in age (age,
age2, and age3) and BMI
(linear). The apo A-1 and B-100 changes in response to exercise
training were adjusted for the effects of polynomial in age and
baseline apo A-1 and B-100 (linear), respectively, separately within
each of the 4 sex-by-generation groups. Each of the adjusted
phenotypes used in the genetic analysis was finally
standardized to a mean of zero and an SD of 1 within each of the 4
sex-by-generation groups.
Commingling
The method of commingling analysis as described in
MacLean et al32 and implemented in the computer program
SKUMIX33 was used. A mixture of up to 3 distributions in
Hardy-Weinberg proportions can be fitted optionally, including p, a
Box-Cox power transformation parameter. There are 5
additional parameters in the model, including the common
variance in each component (E); the overall mean (u) in the entire
sample; q, which determines the relative proportion
(q2) of component distribution with the highest
mean; the displacement between the 2 extreme component means (t); and
the relative position of the mean of the middle component (d). These
parameters were estimated by the maximum-likelihood method.
Hypothesis tests for nested models were carried out with the likelihood
ratio test. The test criterion, which is the difference in minus twice
the log-likelihoods (-2 ln L) obtained under 2 models, is distributed
asymptotically as a 
2, with the degrees of
freedom being equal to the difference in the number of
parameters estimated in the competing models. Nonnested
models are compared by Akaike’s information criterion (AIC), which is
-2 ln L plus twice the number of estimated parameters, and
the "best" model is the one with the smallest AIC.34
Segregation Model
Segregation analysis, as implemented in the computer
program POINTER,32 33 34 35 was performed with the unified
mixed model.36 This model assumes that a phenotype
is composed of the independent and additive contributions from a major
effect, a heritable multifactorial background, and a unique
environmental residual. The major effect is assumed to result from the
segregation at a single locus with 2 alleles (A and
a). The a allele is defined as associated
with higher trait value. Included in the model are 7
parameters: the overall variance (V); the overall mean (u);
the frequency of the a allele (q); the displacement
between the 2 homozygous means (t); the relative position of the
heterozygous mean, or dominance (d); and the multifactorial
heritability in offspring (H) and in parents (HZ). In this study, we
used conditional likelihood (ie, conditioned the likelihood on the
parental values). We note that u, q, t, and d are
parameterized identically to those in SKUMIX. The
transmission pattern of the major gene from parents to offspring is
characterized by 3 parameters: 
1
is the probability that an AA individual transmits
allele A to the offspring; 2 is
the probability that Aa transmits A; and
3 is the probability that aa
transmits A. Under Mendelian transmission,
1=1, 2=0.5, and
3=0, and when the 3 values are equal, no
transmission of the major effect is obtained. The following 3
conditions are usually necessary to infer a major
gene35 : (1) rejection of the no-major-effect
hypothesis (q=t=d=0); (2) nonrejection of the Mendelian transmission
hypothesis (Mendelian ’s); and (3) rejection of the
no-transmission hypothesis (equal ’s). Competing models are tested
for significance by the likelihood ratio test.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
In the present study, reproducibility of baseline apo A-1 and B-100 measures is very high, with coefficients of variation for repeated measures being 4.9% and 6.2% for baseline apo A-1 and B-100, respectively, and intraclass correlations for repeated measures being 0.89 and 0.95 for baseline apo A-1 and B-100, respectively.31 Means and SD of baseline and apo A-1 and B-100 changes in response to exercise training are given in Table 1
. Group differences may be judged by use of the SEM
comparisons. There were significant generation differences in the means
of both baseline apo A-1 and B-100, with higher levels in parents than
in offspring within sex, but not in men for mean of baseline apo A-1
levels. In each of the generations, means of baseline apo A-1 were
significantly higher in women than in men, whereas means of baseline
apo B-100 were higher in men than in women (not significant in
offspring). For apo A-1 and B-100 changes in response to exercise
training, there were no generation differences in the means for men or
women and no sex differences in the means for parents or offspring,
except that mean apo B-100 changes in response to exercise training
were significantly higher in mothers than in fathers.
For age-adjusted baseline apo A-1, there were no significant age
effects in mothers and offspring. For age-adjusted baseline apo B-100,
age was not a significant predictor in parents or daughters. For
age-BMI–adjusted baseline apo A-1 and B-100, BMI accounted for a
larger percentage of variance than did age. For age- and
baseline-adjusted apo A-1 and apo B-100 changes in response to exercise
training, baseline apo A-1 was a significant predictor in parents and
sons, and baseline apo B-100 was a more important predictor in
offspring than in parents. See Table 2
for detailed percentages of variance accounted for by the effects of
age, BMI, and the baseline apo A-1 and B-100 levels.
|
The commingling analysis results suggested that the best models consist of 2 normal distributions for age-adjusted baseline apo A-1 (E=0.99436, u=0.02302, d=0, t=3.71293, q=0.08008, P=1), age-BMI–adjusted baseline apo A-1 (E=0.98692, u=0.02389, d=0, t=3.77115, q=0.08229, P=1), and age-baseline–adjusted apo A-1 change in response to exercise training (E=0.99613, u=0.01238, d=0, t=3.26704, q=0.10320, P=1); 3 skewed distributions for age-adjusted baseline apo B-100 (E=0.37407, u=0.10469, d=0.43438, t=3.58885, q=0.16647, P=2.21982) and age-BMI–adjusted baseline apo B-100 (E=0.43936, u=0.07154, d=0.44020, t=3.02684, q=0.19375, P=1.86593); and 2 skewed distributions for age-baseline–adjusted apo B-100 change in response to exercise training (E=0.76250, u=0.05458, d=0, t=2.51142, q=0.22426, P=1.71526). The finding of multiple distributions is compatible with a major-gene hypothesis; however, commingling may also arise through other causes. Thus, segregation analysis was used to determine whether these major effects segregated in families according to Mendelian expectations.
Segregation analysis results are summarized in Tables 3 and 4 for
baseline and apo A-1 and B changes in response to exercise training,
respectively. The parameter estimates under each of the
most parsimonious segregation models are given in Table 5. For age-adjusted baseline apo A-1,
hypothesis tests for models 2, 3, and 4 were significant, suggesting
that there were significant multifactorial, major, and familial
effects. Both additive (model 7) and dominant (model 8) hypotheses were
rejected, whereas the recessive (model 6) mode of inheritance fit the
data. Tests of transmission probabilities were carried out under the
parsimonious Mendelian hypothesis (no generation difference in the
multifactorial component and a recessive mode of inheritance for the
major-gene component). Although there was a major effect, neither the
hypothesis of Mendelian ’s (model 12, free ’s, "most
general" model in Tables 3 and 4) (model 9-model 12:
32=0.46, P=0.928)
nor the hypothesis of equal ’s (model 11, ’s=1-q,
"nontransmitted" model in Tables 3 and 4) (model
11-model 12: 32=0.09,
P=0.993) was rejected. According to AIC, the most
parsimonious hypothesis was the equal ’s model (AIC=884.38). Thus,
the transmission of the major effect was ambiguous. The major effect
accounted for 9% of the variance, and an additional 56% of the
variance was due to a multifactorial effect. Similar results were
obtained for the age-BMI–adjusted apo A-1: the major effect accounted
for 8% of the variance, and an additional 51% of the variance was due
to a multifactorial effect.
|
|
|
For age-adjusted baseline apo B-100, models 2 to 4 were rejected,
suggesting that there was familial resemblance (both major and a
multifactorial component). The recessive hypothesis (model 6) was
accepted, whereas both additive and dominant hypotheses (models 7 and
8) were rejected. Transmission probability tests were performed under
the parsimonious Mendelian hypothesis (model 9, no generation
difference in the multifactorial component and a recessive mode of
inheritance for the major-gene component), and the 3 required
conditions to infer a major-gene hypothesis were satisfied: First,
major effect was significant (model 3-model 1:
32=33.57,
P<0.001); second, the hypothesis of Mendelian ’s was
not rejected (model 9-model 12:
32=0.23, P=0.891);
finally, the equal ’s hypothesis was rejected (model 11-model 12:
32=8.30, P=0.040).
The putative homogeneous major gene accounted for 56% of
the phenotypic variance, and 25% (q2) of the
sample carried the homozygous recessive genotype. An additional
25% of the variance was due to a multifactorial effect. For the
age-BMI–adjusted baseline apo B-100, a similar pattern resulted,
except that the restricted-equal-’s
(1=2=3=1-q)
hypothesis (model 11) was borderline
(32=7.47,
P=0.058). The putative major gene accounted for 50% of the
phenotypic variance, and 21% (q2) of the sample
carried the homozygous recessive genotype. An additional 29%
of the variance was due to a multifactorial effect.
For age-baseline–adjusted apo A-1 change in response to exercise
training, each of the multifactorial, major, and familial effects was
suggested. Transmission probability tests were carried out under the
parsimonious Mendelian hypothesis (generation differences in the
multifactorial component and a recessive major-locus component).
Neither the Mendelian ’s nor the equal ’s hypothesis was
rejected, suggesting an ambiguous transmission of the major effect from
parents to offspring. Although the AIC suggests that the Mendelian
hypothesis fits best (AIC=835.59), evidence obtained was not clear
enough to infer a major-locus effect.35 The major effect
accounted for 15% of the phenotypic variance, with an additional 21%
and 58% of the variance due to a multifactorial effect in parents and
offspring, respectively. For the age-baseline–adjusted apo B-100
change in response to exercise training, neither the major effect
(model 3) nor the multifactorial component (model 2) was significant,
but there was evidence for familial resemblance (model 4). At least 1
of these effects is needed. The AIC suggests that the best model is for
a multifactorial-only effect (model 5). As given in Table 5,
27% of the variance was due to a multifactorial effect, and no major
effect was found.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The primary purpose of this investigation was to understand whether the segregation patterns of baseline apo A-1 and B-100 in these sedentary white families, as well as their responses to a 20-week endurance exercise training program, supported a major-gene hypothesis. The present investigation represents the first study to assess the major-locus hypothesis for apo A-1 and B-100 changes in response to exercise training.
In this study, correlation coefficients for baseline apo A-1 and B-100
were 
0.35 (P<0.05) in parent-offspring and siblings, in
contrast to 0.10 (P>0.05) in spouses, which suggested
the presence of familiality. They were somewhat lower for the response
to training measures, which were 0.15 to 0.30 in parent-offspring
and siblings.
Varied results from previous segregation studies of apo A-1 have been reported within the last 2 decades. Studies in a large pedigree with excess coronary heart disease18 and in a large pedigree ascertained through cases of early myocardial infarction23 found no evidence of a major gene and even no genetic transmission of apo A-1, whereas findings from other studies supported the hypothesis of a major9 13 14 22 dominant10 11 21 or recessive12 gene, which accounted for 27% to 58% of the phenotypic variance, with an additional component resulting from a polygenic or multifactorial effect (up to 65% variance). Among these studies, Moll et al9 used 283 pedigrees that were randomly selected from the population with respect to disease status and risk factors for coronary artery disease, identified heterogeneous etiologies, and found evidence of a polygenic effect versus a single locus effect from 2 subsets (126 pedigrees supporting nontransmitted environmental factor versus 157 pedigrees supporting genetic etiology) of their data, respectively. Moreover, in other studies, a major-locus hypothesis was not conclusive, unless it was paired with HDL-C10 or a second major locus was involved.22 Recently, Livshits et al15 found a major locus influencing plasma apo A-1 levels also influencing plasma HDL3-C concentrations in a random sample of 228 Israeli pedigrees; however, Juo et al16 used 137 selected families ascertained through probands undergoing elective diagnostic coronary angiography at the Johns Hopkins Hospital and reported no common major locus for these phenotypes. In addition, 2 recent studies used the GAW8 Berkeley data set,21 and the 137 selected families from the Johns Hopkins Hospital22 have provided evidence of 2 major loci affecting apo A-1 levels.
Findings from the present study for age-adjusted and age-BMI–adjusted baseline apo A-1 provided no evidence for a major locus controlling baseline apo A-1 levels, although there was a noteworthy pattern of a major effect with ambiguous transmission from parents to offspring. The major effect accounted for 8% to 9% of the variance and was accompanied by a substantial multifactorial heritability (51% to 56%). These results are similar to the findings from the first data subset (ie, 126 population-based pedigrees supporting the nontransmitted environmental factor) by Moll et al9 and are quite comparable to findings from a single locus model by Juo et al.22 Because BMI adjustments of the data did not significantly affect the results, a pleiotropic effect (ie, a single common gene underlying both baseline apo A-1 and BMI) may not be operating. The non-Mendelian nature of the major effect could be due to environmental factors, or it may be the result of gene-environment interactions that were not modeled. In fact, a second major gene affecting apo A-1 levels, as suggested by a previous investigation,22 was not modeled in the present study and offers 1 possible explanation for the non-Mendelian pattern found here. Other environmental factors potentially affecting levels of apo A-1 include smoking, alcohol consumption, use of exogenous sex hormones, and possible gene-environment interactions that were not considered in this study. In addition, results for the age-baseline–adjusted apo A-1 change in response to training provided no clear evidence to infer a major locus. The major effect, whose transmission from parents to offspring was ambiguous (genetic in nature or environmental in origin), accounted for 15% of the adjusted variance. An additional 21% and 58% of the phenotypic variance was due to a multifactorial component in parents and offspring, respectively. Again, in addition to potential factors discussed above, limited sample size might also be an explanation that prevented the second distribution component from being detectable in this study.
For apo B-100, studies in a large Amish pedigree ascertained through a 13-year-old boy who died suddenly of advanced coronary atherosclerosis24 and in a sample of 102 families volunteered for a free health checkup in the Preventive Center of Vandoeuvre-lès-Nancy of France25 found no evidence of a major-gene effect, but a multifactorial component was significant. In contrast, 6 studies reported positive findings of a major gene influencing apo B-100 levels.2 12 17 18 19 20 Among the findings from these 6 investigations, studies in a large pedigree ascertained through cases of early myocardial infarction17 in a large pedigree with excess coronary heart disease18 in 83 pedigrees ascertained though probands identified by the Lipid Clinic of the Clinical Research Institute of Montreal19 and in 40 Dutch families ascertained through familial combined hyperlipidemia probands20 suggested a codominant major gene that accounted for 35% to 66% of the phenotypic variance. An additional 8% of variance was due to a multifactorial component.19 A study in a random sample of 367 population-based Israeli pedigrees suggested a recessive major gene that accounted for 32% of the variance, with a polygenic effect accounting for an additional 26% of the variance.12 The other study, in 263 families ascertained through children with a family history of coronary artery disease at the Children’s Hospital of Philadelphia, suggested that the mode of inheritance was unresolved between codominant and recessive, with a major locus accounting for 79% of the total variance.2
Results from the present study for age-adjusted and age-BMI–adjusted baseline apo B-100 supported the hypothesis of a major recessive locus regulating baseline apo B-100 levels in these sedentary families. The putative locus accounted for 50% to 56% of the phenotypic variance, and the multifactorial component accounted for an additional 25% to 29% of the variance. Our results did not support a codominant hypothesis predominant in previous reports but were consistent with findings by Livshits et al.12 The recessive major locus accounted for a higher percentage of the phenotypic variance in the present study (50% to 56% in the present study versus 32% by Livshits et al12 ), and there was a similar effect due to a multifactorial component (25% to 29% in the present study versus 26%). However, results for the age-baseline–adjusted apo B-100 change in response to exercise training suggested a multifactorial component and failed to demonstrate a major-gene effect. Thus, the major gene may be specific to baseline levels, and it is not clear whether the same gene also regulates response to training. The percentage accounted for by the multifactorial component for the apo B-100 change in response to exercise training was 27%. Again, there are no available reports of familial basis of apo B-100 change in response to exercise training.
In summary, evidence of a major gene with a recessive mode of inheritance was found for baseline apo B-100 that accounted for 56% of the phenotypic variance in a sample of 99 white families of the HERITAGE Family Study. In addition, a major effect with a multifactorial component for baseline apo A-1 and its response to training and a multifactorial effect for apo B-100 response to training were detected. These results are generally in accordance with findings from previous studies, suggesting that the factors underlying the phenotype levels in these sedentary families are similar to those in the general population. In particular, this is the first study that documented familial basis of apo A-1 and B-100 changes in response to endurance exercise training.
|
Acknowledgments |
|---|
The HERITAGE Family Study is supported by the National Heart, Lung, and Blood Institute through grants HL-45670 (C. Bouchard, PI), HL-47323 (A.S. Leon, PI), HL-47317 (D.C. Rao, PI), HL-47327 (J.S. Skinner, PI), and HL-47321 (J.H. Wilmore, PI) and an NIH grant to the University of Minnesota Clinical Research Center. C. Bouchard is also supported by the MRC of Canada Roche Canada Donald B. Brown Research Chair on Obesity. A.S. Leon also is supported in part by the Henry L. Taylor Professorship in Exercise Science and Health Enhancement.
Received April 5, 1999; accepted September 22, 1999.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
1. Zannis VI, Breslow JL. Genetic mutations affecting human lipoprotein metabolism. In: Harris H, Hirschhorn K, eds. Advances in Human Genetics. New York, NY: Plenum Press; 1984;14:125–215.
2. Jarvik GP, Beaty TH, Gallagher PR, Coates PM, Cortner JA. Genotype at a major locus with large effects on apolipoprotein B levels predicts familial combined hyperlipidemia. Genet Epidemiol. 1993;10:257–70.[Medline] [Order article via Infotrieve]
3. Sing CF, Moll PP. Genetics of atherosclerosis. Annu Rev Genet. 1990;24:171–187.[Medline] [Order article via Infotrieve]
4. Coresh J, Beaty TH, Kwiterovich PO, Antonarakis SE. Inheritance of plasma apolipoprotein B levels in families of patients undergoing coronary arteriography at an early age. Genet Epidemiol. 1993;10:159–176.[Medline] [Order article via Infotrieve]
5. Berg K. The genetics of the hyperlipidemias and coronary artery disease. Proc Clin Biol Res. 1982;103:111–125.
6.
Kuusi T, Kesäniemi A, Vuoristo M, Miettinen TA,
Koskenvuo M. Inheritance of high density lipoprotein and lipoprotein
lipase and hepatic lipase activity.
Arteriosclerosis. 1987;7:421–425.
7. Hamsten A, Iselius L, Dahlée 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]
8. Moll PP, Sing CF, Williams RR, Mao SJT, Kottke BA. The genetic determination of plasma apolipoprotein A-1 levels measured by radioimmunoassay: a study of high-risk pedigrees. Am J Hum Genet. 1986;38:361–372.[Medline] [Order article via Infotrieve]
9. Moll PP, Michels VV, Weidman WH, Kottkell BA. Genetic determination of plasma apolipoprotein A1 in a population-based sample. Am J Hum Genet. 1989;44:124–139.[Medline] [Order article via Infotrieve]
10. Borecki IB, Laskarzewski P, Rao DC. Genetic factors influencing apolipoprotein AI and AII levels in a kindred with premature coronary heart disease. Genet Epidemiol. 1988;5:393–406.[Medline] [Order article via Infotrieve]
11. Prenger VL, Beaty TH, Kwiterovich PO. Genetic determination of high-density lipoprotein-cholesterol and apolipoprotein A-1 plasma levels in a family study of cardiac catheterization patterns. Am J Hum Genet. 1992;51:1047–1057.[Medline] [Order article via Infotrieve]
12. Livshits G, Blettner M, Graff E, Hoting I, Wahrendorf J, Brunner D, Schettler G. Tel Aviv-Heidelberg three-generation offspring study: genetic determinants of apolipoprotein A1 and apolipoprotein B. Am J Med Genet. 1995;57:410–416.
13. Blangero J, Williams-Blangero S, Mahaney MC. Multivariate genetic analysis of apo A-1 concentration and HDL subfractions: evidence for major locus pleiotropy. Genet Epidemiol. 1993;10:617–622.[Medline] [Order article via Infotrieve]
14.
Blangero J, Williams-Blangero S, Mahaney MC, Comuzzie
AG, Hixson JE, Samollow PB, Sharp RM, Stern MP, MacCluer JW. Effects of
a major gene for apolipoprotein A-I concentration are thyroid hormone
dependent in Mexican Americans. Arterioscler Thromb Vasc
Biol. 1996;16:1177–1183.
15. Livshits G, Vainder M, Blettner M, Graff E, Wahendorf J, Brunner D. Major locus influencing plasma apo-A1 levels also controls plasma HDL3-C concentrations. Genet Epidemiol. 1998;15:237–249.[Medline] [Order article via Infotrieve]
16. Juo SHH, Beaty TH, Duffy DL, Coresh J, Kwiterovich PO Jr. No common major gene for apolipoprotein A-1 and HDL3-C levels: evidence from bivariate segregation analysis. Genet Epidemiol. 1999;16:54–68.[Medline] [Order article via Infotrieve]
17. Hasstedt SJ, Wu L, Williams RR. Major locus inheritance of apolipoprotein B in Utah pedigrees. Genet Epidemiol. 1987;4:67–76.[Medline] [Order article via Infotrieve]
18. Amos CI, Elston RC, Srinivasan SR, Wilson AF, Cresanta JL, Ward LJ, Beerenson GS. Linkage and segregation analyses of apolipoprotein A1 and B, and lipoprotein cholesterol levels in a large pedigree with excess coronary heart disease: the Bogalusa heart study. Genet Epidemiol. 1987;4:115–128.[Medline] [Order article via Infotrieve]
19. Pairitz G, Davignon J, Mailloux H, Sing CF. Sources of interindividual variation in the quantitative levels of apolipoprotein B in pedigrees ascertained through a lipid clinic. Am J Hum Genet. 1988;43:311–321.[Medline] [Order article via Infotrieve]
20.
Bredie SJH, van Drongelen J, Kiemeney LA, Demacker PNM,
Beaty TH, Stalenhoef AH. Segregation analysis of plasma
apolipoprotein B levels in familial combined
hyperlipidemia. Arterioscler Thromb Vasc
Biol. 1997;17:834–840.
21. Xu JF, Prenger VL, Coresh J, Beaty TH. Two-locus and bivariate segregation analysis of HDL-C and Apo A-1. Genet Epidemiol. 1993;10:605–610.[Medline] [Order article via Infotrieve]
22. Juo SH, Beaty TH, Xu J, Prenger VL, Coresh J, Kwiterovich PO. Segregation analysis of two-locus models regulating apolipoprotein-A1 levels. Genet Epidemiol. 1998;15:73–86.[Medline] [Order article via Infotrieve]
23. Hasstedt SJ, Albers JJ, Cheung MC, Jorde LB, Wilson DE, Edwards CQ, Cannon WN, Ash KO, Williams RR. The inheritance of high density lipoprotein cholesterol and apolipoproteins A-I and A-II. Atherosclerosis. 1984;51:21–29.[Medline] [Order article via Infotrieve]
24. Beaty TH, Kwiterovich PO, Khoury MJ, White S, Bachorik PS, Smith HH, Teng B, Sniderman A. Genetic analysis of plasma sitosterol, apolipoprotein B, and lipoproteins in a large Amish pedigree with sitosterolemia. Am J Hum Genet. 1986;38:492–504.[Medline] [Order article via Infotrieve]
25. Tiret L, Steinmetz J, Herbeth B, Visvikis S, Rakotovao R, Ducimetiere P, Cambien F. Familial resemblance of plasma apolipoprotein B: the Nancy Study. Genet Epidemiol. 1990;7:187–197.[Medline] [Order article via Infotrieve]
26. Berg A, Halle M, Franz I, Keul J. Physical activity and lipoprotein metabolism: epidemiological evidence and clinical trials. Eur J Med Res. 1997;2:259–264.[Medline] [Order article via Infotrieve]
27. Thompson PD, Yurgalevitch SM, Flynn MM, Zmuda JM, Spannaus Martin D, Saritelli A, Bausserman L, Herbert PN. Effect of prolonged exercise training without weight loss on high-density lipoprotein metabolism in overweight men. Metabolism. 1997;46:217–223.[Medline] [Order article via Infotrieve]
28. Zmuda JM, Yurgalevitch SM, Flynn MM, Bausserman LL, Saratelli A, Spannaus Martin DJ, Haerbert PN, Thompson PD. Exercise training has little effect on HDL levels and metabolism in men with initially low HDL cholesterol. Atherosclerosis. 1998;137:215–221.[Medline] [Order article via Infotrieve]
29. Bouchard C, Leon AS, Rao DC, Skinner JS, Wilmore JH, Gagnon J. The HERITAGE Family Study: aims, design and measurement protocol. Med Sci Sports Exerc. 1995;27:721–729.[Medline] [Order article via Infotrieve]
30. Laurell CB. Quantitative estimation of proteins by electrophoresis in agarose gel containing antibodies. Anal Biochem. 1966;15:45–52.[Medline] [Order article via Infotrieve]
31. Després JP, Gagnon J, Bergeron J, Couillard C, El-Moalem HE, Daw WE, Leon AS, Rao DC, Skinner JS, Wilmore JH, Bouchard C. Plasma post-heparin lipase activities in the HERITAGE Family Study: the reproducibility, gender differences and associations with lipoprotein levels. Clin Biochem. 1999;32:157–165.[Medline] [Order article via Infotrieve]
32. MacLean CJ, Morton NE, Elston RC, Yee S. Skewness in commingled distributions. Biometrics. 1976;32:695–699.[Medline] [Order article via Infotrieve]
33. Morton NE, Rao DC, Lalouel JM. Methods in Genetic Epidemiology. New York, NY: Karger; 1983.
34. Akaike H. A new look at the statistical model identification. IEEE Trans Automat Control. 1974;19:716–723.
35. Lalouel JM, Rao DC, Morton NE, Elston RC. A unified model for complex segregation analysis. Am J Hum Genet. 1983;35:816–826.[Medline] [Order article via Infotrieve]
36. Lalouel JM, Morton NE. Complex segregation analysis with pointers. Hum Hered. 1981;31:312–321.[Medline] [Order article via Infotrieve]
This article has been cited by other articles:
 |
|
 |
|
  C Cote, P Pibarot, J-P Despres, D Mohty, A Cartier, B J Arsenault, C Couture, and P Mathieu Association between circulating oxidised low-density lipoprotein and fibrocalcific remodelling of the aortic valve in aortic stenosis Heart, September 1, 2008; 94(9): 1175 - 1180. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J.-F. Mauger, J. Levesque, M.-E. Paradis, N. Bergeron, A. Tchernof, P. Couture, and B. Lamarche Intravascular Kinetics of C-Reactive Protein and Their Relationships with Features of the Metabolic Syndrome J. Clin. Endocrinol. Metab., August 1, 2008; 93(8): 3158 - 3164. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J.-F. Mauger, P. Couture, N. Bergeron, and B. Lamarche Apolipoprotein C-III isoforms: kinetics and relative implication in lipid metabolism J. Lipid Res., June 1, 2006; 47(6): 1212 - 1218. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Desroches, M.-E. Paradis, M. Perusse, W. R. Archer, J. Bergeron, P. Couture, N. Bergeron, and B. Lamarche Apolipoprotein A-I, A-II, and VLDL-B-100 metabolism in men: comparison of a low-fat diet and a high-monounsaturated fatty acid diet J. Lipid Res., December 1, 2004; 45(12): 2331 - 2338. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||