© 1996 American Heart Association, Inc.
Articles |
Genetic Analysis of the IRS
Pleiotropic Effects of Genes Influencing Insulin Levels on Lipoprotein and Obesity Measures
From the Department of Genetics, Southwest Foundation for Biomedical Research, San Antonio, Tex (B.D.M., C.M.K., M.C.M., J.B., A.G.C., L.D.A., J.W.M.), and the Department of Medicine/Epidemiology, University of Texas Health Science Center, San Antonio (S.M.H., M.P.S.).
Correspondence to Braxton D. Mitchell, PhD, Department of Genetics, Southwest Foundation for Biomedical Research, PO Box 28147, San Antonio, TX 78228-0147. E-mail bmitchel@darwin.sfbr.org.
 |
Abstract |
|---|
 Top Abstract IntroductionMethods Results Discussion References |
|---|
Abstract Insulin resistance is part of a metabolic syndrome that also includes non–insulin-dependent diabetes mellitus, dyslipidemia, obesity, and hypertension. It has been hypothesized that insulin resistance represents the primary physiological defect underlying this syndrome. Since insulin resistance is at least partially genetically determined, we hypothesized that genes influencing insulin resistance would have pleiotropic effects on a number of other traits, including triglyceride (TG) and HDL cholesterol levels, body mass index (BMI) and body fat distribution, and blood pressure levels. To investigate this hypothesis, we analyzed data obtained from individuals in 41 families enrolled in the San Antonio Family Heart Study. Statistical methods that take advantage of the relatedness among individuals were used to differentiate between genetic and nongenetic (ie, environmental) contributions to phenotypic variation between traits. Serum levels of fasting and 2-hour insulin (measured in 767 and 743 nondiabetic family members, respectively) were used as a measure of insulin resistance. The genetic correlations were high between insulin levels (both fasting and 2-hour) and each of the following: BMI, HDL level, waist-to-hip ratio, and subscapular-to-triceps ratio, indicating that the same gene, or set of genes, influences each pair of traits. In contrast, the genetic correlations of insulin levels with systolic and diastolic blood pressures were low. We have previously shown that a single diallelic locus accounts for 31% of the phenotypic variation in 2-hour insulin levels in this population. We conducted a bivariate segregation analysis to see if the common genetic effects on insulin and these other traits could be attributable to this single locus. These results indicated a significant effect of the 2-hour insulin locus on fasting insulin levels (P=.02) and BMI (P=.05), with the "high" insulin allele associated with higher levels of fasting insulin but lower levels of BMI. There was no detectable effect of this locus on HDL level, TG level, subscapular-to-triceps ratio, or blood pressure. Overall, these results suggest that a common set of genes influencing insulin levels also influences other insulin resistance syndrome–related traits, although for the most part this pleiotropy is not attributable to the 2-hour insulin level major locus.
Key Words: insulin • body mass index • genetics • pleiotropy
|
Introduction |
|---|
|
TopAbstractIntroductionMethods Results Discussion References |
|---|
Non–insulin-dependent diabetes mellitus (NIDDM), hyperlipidemia, obesity, and hypertension are important contributors to cardiovascular risk. Moreover, the risk associated with these disorders is accentuated because these conditions cluster together more often than would be expected by chance alone. It is hypothesized that these disorders may form part of a metabolic syndrome and that insulin resistance could be the primary physiological defect underlying this syndrome.1 Support for this hypothesis has been provided by the fact that measures of insulin resistance are correlated with these disorders both in cross-sectional2 3 4 5 6 and prospective7 8 studies.
The underlying cause of insulin resistance is not known, although it is speculated that the primary defect may be genetic.9 If so, and if this defect in turn accounted for the other metabolic abnormalities associated with insulin resistance, then the high correlations consistently observed between measures of insulin resistance and other metabolic disturbances could be attributable to a common genetic pathway or set of pathways. Pedigree studies provide one means for disentangling genetic and environmental (nongenetic) sources of covariation among traits.
In this study we investigated the hypothesis that the genes that influence insulin sensitivity, as measured by insulin levels,10 exert pleiotropic effects on other insulin resistance syndrome (IRS)–related traits, including obesity and body fat distribution, triglyceride and HDL cholesterol levels, and blood pressure levels. To accomplish these goals, statistical genetic methods were used to analyze data on a large group of related individuals who were examined as part of the San Antonio Family Heart Study (SAFHS). These methods take advantage of the relatedness among individuals to differentiate between genetic and nongenetic (ie, environmental) contributions to phenotypic variation among traits.
|
Methods |
|---|
|
TopAbstractIntroductionMethods Results Discussion References |
|---|
The SAFHS was designed to investigate the genetics of heart disease and its determinants among Mexican Americans in San Antonio, Tex. Family members were enrolled into the SAFHS on the basis of a proband who was randomly ascertained from a low-income neighborhood and who was between the ages of 40 and 60 years. The probands were identified without regard to disease status. Probands and all first-, second-, and third-degree relatives of the probands aged 16 years and older were invited to attend our clinic to receive a medical examination.
Between January 1992 and August 1994, a total of 921 individuals were enrolled into the SAFHS. These individuals comprised 41 different pedigrees, ranging in size from 3 to 76 individuals. The ages of participating family members ranged from 16 to 92 years. The sample included information on 1748 pairs of first-degree relatives (838 parent-offspring pairs and 910 sibling pairs), 2262 pairs of second-degree relatives (1897 avuncular pairs, 270 grandparent-grandchild pairs, and 95 half-sibling pairs), and 2414 pairs of third-degree relatives (1734 cousin pairs, 405 great-avuncular pairs, 257 half-avuncular pairs, and 18 great grandparent–great grandchild pairs).
Clinical Measurements
Participating individuals attended our
medical clinic in the
mornings after an overnight (12-hour) fast. A fasting blood sample was
obtained for determination of glucose, insulin, and lipid and
lipoprotein levels. Cholesterol and
triglyceride concentrations were assayed enzymatically on
frozen plasma samples with the use of a Gilford SBA-300 clinical
chemistry analyzer with commercial reagents supplied by
Boehringer-Mannheim Diagnostics and Stanbio,
respectively. HDL cholesterol was measured after
precipitation of apolipoprotein B–containing particles from freshly
collected plasma by the use of dextran sulfate.11 Plasma
glucose was measured using an Abbott V/P analyzer, and serum
insulin concentrations were measured using a commercial
radioimmunoassay kit (Diagnostic Products). Subjects
then ingested a 75-g glucose equivalent load, and glucose and insulin
levels were again determined from a blood sample obtained 2 hours after
the glucose challenge. Diabetes was diagnosed according to the World
Health Organization plasma glucose criteria.12 Subjects
currently taking antidiabetic medications were also considered to have
diabetes.
The systolic (first phase) and diastolic (fifth phase) blood pressures were measured to the nearest even digit using a random-zero sphygmomanometer (Hawksley-Gelman) on the right arm of the seated participant. Three readings were recorded for each individual, and the subject's blood pressure was defined as the average of the second and third readings.
Weight and height were also obtained, and body mass index was calculated as weight (kilograms) divided by height (meters) squared. Adiposity measures were also obtained at several local sites. Waist and hip circumferences were measured with a steel tape at the level of the umbilicus and the greater trochantors, respectively, and skinfold thicknesses were measured with Lange calipers.13
For purposes of these analyses, triglyceride and HDL cholesterol levels were considered missing from 15 individuals who reported that they were currently taking lipid-lowering medications and from an additional 26 individuals reporting a previous history of heart attack or heart surgery. Similarly, blood pressure levels were excluded from 80 individuals reporting current usage of antihypertensive medications. Because diabetes may cause secondary changes in insulin levels, diabetic individuals (n=128) were excluded from all analyses involving insulin levels. Thus, of 921 examined individuals, lipid levels were potentially available on 880, blood pressures on 841, and insulin levels on 793. Prior to analysis, insulin and triglyceride levels were transformed by their natural logarithms to reduce skewness in the data. No other prior adjustments to the data were made.
Statistical Methods
Management of pedigree and phenotype data
was
accomplished using the computer package PEDSYS.14
Statistical genetic analyses were performed using our modified
version of the pedigree analysis program PAP
V3.0,15 which uses maximum likelihood methods to compute
the likelihoods of the pedigree data under different transmission
models.
According to classic quantitative genetic
theory,16 the
total phenotypic variance of a trait,
2P, can be partitioned into
2G, the variance due to the effects
of genes, and 2E, the variance due to
environmental effects. These components are additive, such that
2P=2G+2E.
Each of these components can be further partitioned. The heritability
(h2) of a phenotype corresponds to that proportion
of the total phenotypic variance that is attributable to the additive
effects of genes and is calculated as
2G/2P.
We initially estimated the mean, standard deviation, and residual heritability of each trait, while simultaneously estimating the effects of age, sex, and diabetes status. The results of these analyses were then used as the starting values for the bivariate analyses described below.
All hypotheses concerning pleiotropy were evaluated through the use of bivariate analysis methods. Two classes of hypotheses were considered. The first class involved the use of quantitative bivariate genetic analysis to quantify the pleiotropic effects of additive genes affecting insulin levels and each of seven other IRS-related traits. In other words, results of these analyses enabled us to test the hypothesis that two traits are influenced by the same set of genes. In contrast, the goal of the second class of hypotheses that we considered was to evaluate whether a single locus-specifically a locus that we have previously shown to account for 31% of the total residual phenotypic variability in 2-hour insulin levels17 -had any effect on other IRS-related traits. This latter set of hypotheses was tested using bivariate segregation analysis methods.
Parameters estimated from the bivariate genetic analyses were used to make inferences about pleiotropy. Briefly, univariate genetic analysis (ie, analysis of a single phenotype, with or without covariates) may be extended to the multivariate state (ie, joint analysis of multiple phenotypes) by modeling the multivariate phenotype as a linear function of the individual's phenotypic values, the population means of these traits, the covariates and their regression coefficients, plus the additive genetic values and random environmental deviations.18 19 20 From such a model, and making use of the kinship coefficients among individuals, the genetic and environmental variance-covariance matrices may be obtained, and from these the environmental and genetic correlations among phenotype pairs may be computed.
The genetic and environmental
correlations obtained from such models
represent estimates of the effects of shared genes, or
pleiotropy, and of shared environmental factors, respectively, on the
phenotypic variance in a trait. The additive genetic (
G)
and environmental (E) correlations between two traits,
in turn, may then be used to estimate the total phenotypic correlation
between two traits, (P), as follows:
 |
Thus, the phenotypic correlation may be viewed as a weighted average of the genotypic and environmental correlations, with the weights being a function of the heritabilities of the two traits. Standard errors for the phenotypic correlations may be obtained by means of a first-order Taylor series approximation.
We conducted a
series of bivariate quantitative genetic and segregation
analyses of both insulin measures with each of the other seven
traits using the simultaneous orthogonalization methods of
Blangero and Konigsberg,21 as implemented in our modified
version of PAP V3.0.15 In the quantitative genetic
analyses assessing polygenic pleiotropy, the likelihood of the
pedigree data was maximized under polygenic models that did not allow
for major locus effects. Each bivariate polygenic pleiotropy model
included 20 parameters. For example, to determine if a
common set of genes influenced both insulin levels and blood pressure
levels, estimates were obtained, using maximum likelihood, of the
overall phenotypic mean (µ), standard deviation (), and the
heritability (h2) of each paired phenotype, as well
as the coefficients associated with following covariate effects for
each paired phenotype: sex, sex-specific age effects and
sex-specific age2 effects, and body mass index. The two
additional parameters that were estimated corresponded to
the genetic and environmental correlations between the two traits (see
Table 1
).
|
The hypothesis of polygenic pleiotropy was
evaluated by a likelihood
ratio test, calculated as the difference in -2xln likelihoods
between a restricted model (a model in which the value of the genetic
correlation was fixed at zero) and an unrestricted model (a model in
which all parameters are estimated). The likelihood ratio
statistic is distributed asymptotically as a 
2
statistic with degrees of freedom equal to the difference in number of
parameters in the two models being
compared.22
After evaluation of the polygene pleiotropy
hypotheses, a second set of
analyses was undertaken to evaluate the hypothesis of single
locus pleiotropy. This latter class of models differed from the
bivariate polygene pleiotropy models in that they allowed for an
additional effect of a single diallelic locus, which could influence
both 2-hour insulin levels and the second trait. This extension leads
to the addition of the following parameters to the genetic
model: the frequency of the "high" insulin allele in the
population (pA), three genotype-specific means
for each phenotype (µAA,
µAa, µaa), and the transmission
probabilities (
) that an individual of specified genotype
will transmit allele A to his/her offspring (under
mendelian expectations, AA=1.0,
Aa=0.5,
aa=0). Thus, each major locus bivariate model included
25 parameters to be estimated. Eleven
parameters were required to characterize each trait
individually (three genotypic means and the common standard deviation,
the residual heritability, and six covariate effects), as well as the
frequency of one of the two alleles at the major locus and the
residual genetic and environmental correlations between the two traits
after accounting for the major locus effect. Table 1
shows the
parameters estimated from the bivariate polygene pleiotropy
and bivariate single locus pleiotropy models.
In the polygene pleiotropy analysis, pleiotropy is indicated by additive genetic correlations between two traits that are significantly different from zero as assessed by the likelihood ratio test. In the major locus pleiotropy analysis, major locus pleiotropy is indicated when the likelihood of the bivariate model that includes genotype-specific means for the second trait (ie, genotype at the 2-hour insulin locus influences phenotypic levels of the second trait) is significantly higher than the likelihood of a bivariate model in which only a single phenotypic mean is estimated for the second trait (ie, the 2-hour insulin locus does not influence mean levels of the second trait). A significant residual additive genetic correlation in the major locus pleiotropy model indicates the existence of shared additive genetic effects on the two traits over and above that attributable to the 2-hour insulin locus.
|
Results |
|---|
|
TopAbstractIntroductionMethods Results Discussion References |
|---|
Clinical characteristics of the study population are shown in Table 2
. The mean age was 38.4 years in men and 39.3
years in women. Diabetes was present in 13.6% of men and 14.1% of
women (n=128 individuals). Median fasting insulin levels were 78.8
pmol/L in men and 81.6 pmol/L in women. Median 2-hour insulin levels
were 349.6 pmol/L and 490.3 pmol/L in men and women, respectively.
Women had significantly higher levels of 2-hour insulin, body mass
index, and HDL cholesterol compared with men, while levels
of triglycerides, systolic and
diastolic blood pressures, subscapular-to-triceps
ratio, and waist-to-hip ratio were significantly higher in men.
Fasting insulin levels did not differ significantly between the
sexes.
|
The univariate heritabilities of fasting and 2-hour
insulin, body mass index, subscapular-to-triceps ratio,
waist-to-hip ratio, systolic and diastolic
blood pressure, HDL cholesterol, and
triglycerides are shown in Table 2. After accounting for
the variation due to age, sex, and diabetes status, the traits with the
highest heritabilities were HDL cholesterol (54.3%), body
mass index (53.0%), and triglycerides (49.8%). The
heritabilities of fasting and 2-hour insulin were 44.2% and 20.1%,
respectively. The subscapular-to-triceps ratio had a
heritability of 44.9%, while the heritability of the
waist-to-hip ratio was only 17.3%. The heritabilities for
systolic and diastolic blood pressure were 32.8%
and 41.0%, respectively. The heritability for each of these traits was
significantly greater than zero (P
.001 for all
traits).
Do Insulin Levels and Other IRS-Associated Traits Have a Common
Genetic Background?
Bivariate quantitative genetic analyses were
performed to
evaluate the hypothesis that a common set of genes has pleiotropic
effects on both insulin levels and other insulin
resistance–associated phenotypes. The phenotypic, genetic,
and environmental correlations obtained from these analyses are
shown in Table 3. The largest overall phenotypic
correlations were between insulin levels and body mass index, HDL
cholesterol levels, and triglyceride levels.
Fasting and 2-hour insulin levels were also highly correlated with each
other (Phen=0.448±.044). Genetic and
environmental
factors contributed more or less equally to the correlations between
insulin and body mass index. In contrast, the genetic correlations
between insulin and both HDL cholesterol and
triglycerides were considerably higher than the
corresponding environmental correlations between insulin and these
traits. Similarly, the genetic correlations between insulin levels and
waist-to-hip ratio and between 2-hour insulin levels and
subscapular-to-triceps ratio were substantially higher than the
corresponding environmental correlations between these traits.
|
The
magnitudes of the genetic correlations between traits correspond to
the degree of pleiotropy. The highest genetic correlation observed was
between fasting and 2-hour insulin (G=0.544),
indicating
that common genes influence both traits. The genetic correlations were
moderately high between insulin and HDL cholesterol
(-0.356 and -0.285 for fasting and 2-hour insulin,
respectively), triglycerides (0.297 and 0.400 for fasting
and 2-hour insulin, respectively), and body mass index (0.486 and 0.379
for fasting and 2-hour insulin, respectively). The genetic correlations
between 2-hour insulin levels and both subscapular-to-triceps
ratio (G=0.336) and waist-to-hip ratio
(G=0.339) were also relatively high. In contrast, the
genetic correlations between insulin and both systolic and
diastolic blood pressures were low and did not differ
significantly from zero.
Is There Evidence for a Single Gene Influencing Insulin Levels That
Also has Detectable Effects on Other IRS Traits?
Evidence for a major
gene influencing 2-hour insulin
concentrations in this population has been reported
previously.17 Prior to the bivariate segregation
analyses, model parameters corresponding to this
major gene locus were reestimated from univariate
segregation analysis using the current, expanded data set.
Parameter estimates were similar to those reported
previously. As before, the inheritance of 2-hour insulin levels was
best described using an autosomal dominant transmission model. The
frequency of the allele associated with high insulin levels was
0.668, with individuals homozygous or heterozygous for this allele
having mean insulin levels equal to 3.964 (back-transformed mean,
467.1 pmol/L). Individuals lacking this allele had mean levels of
2.663 (back-transformed mean, 127.2 pmol/L). The residual
heritability in 2-hour insulin levels after accounting for variation at
this major locus was 18.4%. The two-allele locus accounted for
28.4% of the phenotypic variability in 2-hour serum insulin levels
after accounting for the effects of age, sex, and body mass index.
Bivariate segregation analysis then was performed to determine whether this locus exerted an effect on other traits. Parameters corresponding to the inheritance of the 2-hour insulin locus were fixed at their values obtained from the univariate analysis; however, allowing these parameters to be simultaneously estimated along with all other parameters did not appreciably change the results. Under a model allowing for major locus pleiotropy, mean levels of the second trait (eg, fasting insulin or body mass index) were dependent on the individual's putative genotype at the 2-hour insulin major locus. Under a model that does not allow for pleiotropy due to this major locus, mean levels of the second trait did not depend on the major locus genotype (ie, µAA=µAa=µaa).
The results of the bivariate segregation analyses testing
whether the 2-hour insulin locus influences levels of fasting insulin
and body mass index are shown in Table 4. The effect of
the 2-hour insulin locus on fasting insulin and body mass index was
obtained by estimating the mean level of fasting insulin (or body mass
index) associated with each of the 2-hour insulin genotypes.
Because the mean levels of fasting insulin/body mass index associated
with the Aa genotype were not significantly different from
those associated with the AA genotype (ie, a two-mean
[dominant] model fit the data as well as a three-mean
[codominant] model), subsequent tests of major locus pleiotropy were
performed by comparing the dominant models with single mean models,
which did not allow for major locus pleiotropy.
|
For both fasting
insulin and body mass index, the likelihood of the
major locus pleiotropy model was significantly higher than that of the
model constrained to have no major locus pleiotropy
(21=5.40, P=.02 for
fasting insulin; 21=3.79,
P=.05 for body mass index). The A allele, which was
associated with high 2-hour insulin levels, was also associated with
high levels of fasting insulin, with AA and Aa
individuals having mean fasting insulin levels equal to 2.468
(back-transformed mean, 100.0 pmol/L), and aa
individuals having mean levels equal to 2.254 (back-transformed
mean, 80.7 pmol/L). In addition, individuals with the AA or
Aa genotypes had, on average, lower levels of body
mass index than individuals with the aa genotype
(29.17 kg/m2 versus 31.33 kg/m2). Allowing for
the effect of sex to differ across genotypes did not
significantly improve the likelihood of the data. The 2-hour insulin
major locus accounted for 1.4% and 1.2% of the residual phenotypic
variability in fasting insulin levels and body mass index,
respectively. After accounting for this major locus, the residual
genetic correlations remained high between each of these traits and
2-hour insulin levels (G=0.645 between 2-hour and
fasting insulin; and G=0.615 between 2-hour insulin and
body mass index), suggesting that considerable pleiotropy exists
between these traits even after accounting for the pleiotropy
attributable to the 2-hour insulin locus.
There was no detectable effect of the 2-hour insulin locus on levels of triglycerides, HDL cholesterol, subscapular-to-triceps ratio, waist-to-hip ratio, or systolic or diastolic blood pressure (data not shown).
|
Discussion |
|---|
|
TopAbstractIntroductionMethods Results Discussion References |
|---|
The Mexican American population, at least in San Antonio, is characterized by a high degree of insulin resistance as evidenced by the fact that they are relatively hyperinsulinemic compared with non-Hispanic whites.23 Moreover, this relative hyperinsulinemia persists even after accounting for their relative obesity.23 An excess of insulin resistance among Mexican Americans compared with non-Hispanic whites has also been confirmed using the intravenous glucose tolerance test, a more precise measure of insulin resistance.24 25 It has been hypothesized that this insulin resistance may contribute to the adverse cardiovascular risk profile observed in this population, which includes obesity, hypertriglyceridemia and low HDL cholesterol levels, elevated blood pressure levels, and a high incidence of NIDDM.9
There is abundant evidence that insulin resistance has genetic determinants. The nondiabetic first-degree relatives of patients with NIDDM exhibit a range of metabolic defects associated with insulin action.26 27 28 Familial clustering of insulin sensitivity has been reported both in Pima Indians29 and in white populations.30 Significantly, defects in insulin sensitivity among first-degree relatives of diabetic probands occur long before impairment in acute phase insulin release can be detected.30 31
Numerous studies have demonstrated that insulin levels are influenced by genetic variation. Insulin levels are higher in nondiabetic offspring of diabetic parents compared with the nondiabetic offspring of nondiabetic parents.32 In large extended families, Elbein et al33 reported that nondiabetic relatives of diabetic probands had higher insulin levels than control subjects. Formal segregation analyses of insulin levels have further suggested that a single gene may have large effects on insulin levels. In non-Hispanic white families, Schumacher et al34 reported that fasting and 1-hour post–glucose challenge insulin levels are influenced by a single major gene, and more recently, evidence for a major gene influencing 2-hour post–glucose challenge insulin levels has been reported in Mexican Americans.17 Prochazka et al35 reported a sib-pair analysis that provided tentative evidence that a gene influencing insulin sensitivity in Pima Indians was located on chromosome 4q, and subsequently, Mitchell et al36 performed a combined segregation and linkage analysis on chromosome 4q markers in a different set of Mexican American families and obtained a lod score of 2.8 between fatty acid binding protein (FABP2) and the major locus for 2-hour insulin levels at an estimated recombination fraction of 0. However, there was not strong evidence for linkage between FABP2 and the major locus for 2-hour insulin levels in the present population.17
The high heritabilities associated with insulin levels, triglyceride and HDL cholesterol levels, body mass index, subscapular-to-triceps ratio, and blood pressure are consistent with findings from other studies and indicate that genes contribute significantly to phenotypic variability in these traits. In contrast, the waist-to-hip ratio, which is commonly used as a measure of upper-to-lower body fat distribution, has a low heritability in this population. Considerably higher heritability estimates for waist-to-hip ratio have been reported in other populations,37 38 and the relatively low heritability of this measure among Mexican Americans in San Antonio appears to be attributable to the relatively large contribution of environmental (ie, nongenetic) effects to the total phenotypic variance of waist-to-hip ratio in this population.
The high genetic correlations observed among fasting and 2-hour insulin, body mass index, triglyceride and HDL cholesterol levels, and subscapular-to-triceps ratio (correlated with 2-hour insulin levels only) indicate that these traits are influenced by a common gene or set of genes. Moreover, the correlations between insulin and HDL cholesterol and triglycerides were independent of the effects of body mass index, implying that the pleiotropic effects of genes influencing insulin levels involve at least some genes that do not influence obesity. On the other hand, the genetic correlations between insulin (both fasting and 2-hour values) and blood pressure (both systolic and diastolic) were low, and not statistically different from zero, implying that the gene(s) associated with insulin resistance do not have pleiotropic effects on blood pressure.
The bivariate segregation analyses revealed little evidence for major locus pleiotropy between the locus influencing 2-hour insulin levels and other traits associated with the IRS, with the exception that this locus had a significant effect on fasting insulin levels. Specifically, individuals either homozygous or heterozygous for the allele associated with high 2-hour insulin levels also had high fasting insulin levels. This locus, while accounting for 28% of the phenotypic variability in 2-hour insulin levels, accounted for only 1.4% of the variability in fasting insulin levels. These analyses also indicate that other genes contribute substantially to the variability in fasting insulin levels; the residual heritability was approximately 44% even after accounting for the 2-hour insulin major locus. Furthermore, considerable pleiotropy remains between 2-hour and fasting insulin levels, as evidenced by the residual genetic correlation of 0.645 between these two traits.
The bivariate segregation analysis reveals also that the 2-hour
insulin major locus has a small but significant effect on body mass
index, although the direction of this effect is opposite to that
expected from the IRS hypothesis; ie, individuals with the high insulin
allele have lower, not higher, levels of body mass index. After
accounting for age and sex effects, this locus accounted for only a
small portion of the phenotypic variability in body mass index (1.2%).
The positive direction of the genetic correlation between 2-hour
insulin and body mass index (G=0.379) indicates that
the
net pleiotropic effect of all genes influencing insulin and body mass
index is in a positive direction; that is, genes associated with higher
insulin levels are also associated with higher body mass index. This
hypothesis is further supported by the fact that, after accounting for
the 2-hour insulin major locus, the genetic correlation between 2-hour
insulin and body mass index increases from 0.379 to 0.615 (see Table
4). Thus, the 2-hour insulin locus appears to be only one of
multiple
loci influencing both insulin and body mass index levels and has only a
very small effect on body mass index. This small negative effect on
body mass index is probably overwhelmed by other loci with relatively
large effects for which allelic variation leads to increases (or
decreases) in both traits. In fact, we have preliminary evidence for
the existence of a second locus that has a large effect on body mass
index in which the high body mass index allele is associated with
high insulin levels (B. Mitchell, unpublished data, 1995).
On the whole, these analyses indicate the presence of substantial pleiotropy between insulin levels and other traits associated with the IRS, with the one exception being the absence of pleiotropy between insulin and blood pressure. Our analyses provide no evidence, however, that these pleiotropic effects are attributable to the previously identified major locus that influences 2-hour insulin levels. Thus, these analyses provide no evidence for the existence of a single "insulin resistance gene." Obviously, we cannot rule out the possibility that such a gene might exist. For example, it is possible that the parameter estimates obtained from our segregation analysis of 2-hour insulin levels are sufficiently imprecise (perhaps because our major gene model failed to capture sufficiently the underlying complexity of this trait) that we failed to detect a true effect of a locus influencing 2-hour insulin on lipid levels and blood pressure. Other loci influencing insulin levels may also be involved. Despite the fact that the heritability of fasting insulin levels is considerably higher than that of 2-hour insulin levels, we have been unable to detect a major gene for the former. It is certainly possible that a locus with a relatively large effect on fasting insulin does exist (or at least a locus with a larger effect than the 2-hour insulin locus), and possibly this locus has pleiotropic effects on other IRS traits.
The possibility should also be considered that pleiotropic effects of a single "insulin resistance gene" might have been detected had we used a more precise measure of insulin resistance. In our study, serum insulin levels were used as an indirect measure of insulin resistance. Although increased insulin concentrations after a glucose load are strongly associated with insulin resistance in nondiabetic individuals,10 39 insulinemia is an imperfect measure of insulin sensitivity. In general, hyperinsulinemia can be viewed as a compensatory adaptation to insulin resistance, but other factors, including insulin secretory capability,40 also influence circulating insulin levels.
These analyses highlight the complex relationships existing in this Mexican American population between insulin and other traits regarded as components of the IRS. In other populations these relationships may be more, or less, complex. There are some unique features of the Mexican American population in San Antonio. This population is almost entirely of Mexican origin and contains a mixture of both new and old immigrants. It is estimated that the degree of Amerindian genetic admixture in this population ranges from 30% to 50%.41 42 Like several Amerindian tribes of the southwestern United States, the Mexican American population of San Antonio is characterized by a high prevalence of obesity and insulin resistance. In the context of this genetic background, insulin resistance genes may play a more (or less) important role in this population compared with others in influencing phenotypic variation in lipoprotein, blood pressure, and obesity-related traits.
|
Acknowledgments |
|---|
This research was supported by a grant from the National Heart, Lung, and Blood Institute (PO1-HL45522).
Received June 9, 1995; accepted December 1, 1995.
|
References |
|---|
|
TopAbstractIntroductionMethods Results Discussion References |
|---|
1. Reaven GM. Role of insulin resistance in human disease: Banting Lecture 1988. Diabetes. 1988;37:1595-1607. [Abstract]
2. Modan M, Halkin H, Almog S, Luskey A, Eshkol A, Shefi M, Shitrit A, Fuchs Z. Hyperinsulinemia: a link between hypertension, obesity, and glucose intolerance. J Clin Invest. 1985;75:809-817.
3. Zavaroni I, Bonora E, Pagliara M, Dall'Aglio E, Luchetti L, Buonanno G, Bonati PA, Bergonzani M, Gnudi L, Passeri M, Reaven G. Risk factors for coronary artery disease in healthy persons with hyperinsulinemia and normal glucose tolerance. N Engl J Med. 1989;320:702-706. [Abstract]
4. Ferrannini E, Haffner SM, Mitchell BD, Stern MP. Hyperinsulinemia: the key feature of a cardiovascular and metabolic syndrome. Diabetologia. 1991;34:416-422.[Medline] [Order article via Infotrieve]
5.
Cambien F, Warnet J-M, Eschwege E, Jacqueson A,
Richard JL, Rosselin G. Body mass, blood pressure, glucose, and
lipids: does plasma insulin explain their relationships?
Arteriosclerosis. 1987;7:197-202.
6.
Orchard TJ, Becker DJ, Bates M, Kuller LH, Drash
AL. Plasma insulin and lipoprotein concentrations: an
atherogenic association? Am J Epidemiol. 1983;118:326-337.
7. Haffner SM, Valdez RA, Hazuda HP, Mitchell BD, Morales PA, Stern MP. A prospective analysis of the Insulin Resistance Syndrome. Diabetes. 1992;41:715-722. [Abstract]
8.
Mitchell BD, Haffner SM, Hazuda HP, Valdez RA, Stern
MP. The relation between serum insulin levels and 8-year changes
in lipid, lipoprotein, and blood pressure levels. Am J
Epidemiol. 1992;136:12-22.
9. DeFronzo RA, Ferrannini E. Insulin resistance: a multifaceted syndrome responsible for NIDDM, obesity, hypertension, dyslipidemia, and atherosclerotic cardiovascular disease. Diabetes Care. 1991;14:173-194. [Abstract]
10.
Laakso M. How good a marker is insulin level for
insulin resistance? Am J Epidemiol. 1993;137:959-965.
11. Warnick GR, Benderson J, Albers JJ. Dextran sulfate-Mg2+ precipitation procedure for quantification of high-density-lipoprotein cholesterol. Clin Chem. 1983;10:91-99.
12. World Health Organization Expert Committee. Second Report on Diabetes Mellitus. Geneva, Switzerland: World Health Organization; 1980. Technical Report Series No. 646.
13. Haffner SM, Stern MP, Hazuda HP, Pugh JA, Patterson JK, Malina R. Upper body and centralized adiposity in Mexican Americans and non-Hispanic whites: relationship to body mass index and other behavioral and demographic variables. Int J Obesity. 1986;10:493-502. [Medline] [Order article via Infotrieve]
14. Dyke B. PEDSYS: A Pedigree Data Management System. Population Genetics Laboratory. San Antonio, Tex: Southwest Foundation for Biomedical Research; 1994.
15. Hasstedt SJ. Pedigree Analysis Package V 3.0. Salt Lake City, Utah: Department of Genetics, University of Utah; 1989.
16. Falconer DS. Introduction to Quantitative Genetics. 3rd ed. London, England; Longman; 1989.
17. Mitchell BD, Kammerer CM, Hixson JE, Atwood LD, Hackleman S, Blangero J, Haffner SM, Stern MP, MacCluer JW. Evidence for a major gene affecting post-challenge insulin levels in Mexican Americans. Diabetes. 1995;44:284-289. [Abstract]
18. Hopper JL, Mathews JD. Extensions to multivariate normal models for pedigree analysis. Ann Hum Genet. 1982;46:373-383.
19. 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]
20.
Boehnke M, Moll PP, Kottke BA, Weidman WH.
Partitioning the variability of fasting glucose levels in
pedigrees. Am J Epidemiol. 1987;125:679-689.
21. Blangero J, Konigsberg LW. Multivariate segregation analysis using the mixed model. Genet Epidemiol. 1991;8:299-316. [Medline] [Order article via Infotrieve]
22. Edwards AWF. Likelihood. Expanded Edition. Baltimore, Md: Johns Hopkins University Press; 1992.
23. Haffner SM, Stern MP, Hazuda HP, Pugh JA, Patterson JK. Hyperinsulinemia in a population at high risk for non-insulin dependent diabetes mellitus. N Engl J Med. 1986;315:220-224. [Abstract]
24. Haffner SM, Stern MP, Dunn JF, Mobley M, Blackwell J, Bergman RN. Diminished insulin sensitivity, and increased insulin response in non-obese, non-diabetic, Mexican Americans. Metabolism. 1990;39:842-847. [Medline] [Order article via Infotrieve]
25. D'Agostino R Jr, Rewers M, Kades W, Selby J, Savage P, Haffner S, for the IRAS Investigators. Insulin sensitivity, acute insulin response and non-insulin mediated glucose disposal in non-diabetic blacks, Hispanics, and non-Hispanic whites. Diabetes. 1995;44 (suppl 1):5A. Abstract.
26. Erickson J, Franssila-Kallunki A, Ekstrand A, Saloranta C, Widén E, Schalin C, Groop L. Early metabolic defects in persons at increased risk for non-insulin-dependent diabetes mellitus. N Engl J Med. 1989;321:337-343. [Abstract]
27. Gulli G, Ferrannini E, Stern MP, Haffner SM, DeFronzo RA. The metabolic profile of non-insulin-dependent diabetes mellitus is fully established in glucose tolerant offspring of two non-insulin dependent diabetic parents. Diabetes. 1992;41:1575-1586. [Abstract]
28.
Laws A, Stefanick M, Reaven G. Insulin
resistance and hypertriglyceridemia in
nondiabetic relatives of patients with non-insulin-dependent
diabetes mellitus. J Clin Endocrinol
Metab. 1989;69:343-347.
29. Lillioja S, Mott DM, Zawadzki JK, Young AA, Abbott WGH, Knowler WC, Bennett PH, Moll P, Bogardus C. In vivo insulin action is familial characteristic in nondiabetic Pima Indians. Diabetes. 1987;36:1329-1335. [Abstract]
30. Martin BC, Warram JH, Rosner B, Rich SS, Soeldner JS, Krolewski AS. Familial clustering of insulin sensitivity. Diabetes. 1992;41:850-854. [Abstract]
31. Warram JH, Martin BC, Krolewski AS, Soeldner JS, Kahn CR. Slow glucose removal rate and hyperinsulinemia precede the development of type II diabetes in the offspring of diabetic patients. Ann Intern Med. 1990;113:909-915.
32. Haffner SM, Stern MP, Hazuda HP, Mitchell BD, Patterson JK. Increased insulin concentrations in nondiabetic offspring of diabetic parents. N Engl J Med. 1988;319:1297-1301. [Abstract]
33. Elbein SC, Maxwell TM, Schumacher MC. Insulin and glucose levels and prevalence of glucose intolerance in pedigrees with multiple diabetic siblings. Diabetes. 1991;40:1024-1032. [Abstract]
34. Schumacher MC, Hasstedt SJ, Hunt SC, Williams RR, Elbein SC. Major gene effect for insulin levels in familial NIDDM pedigrees. Diabetes. 1992;41:416-423. [Abstract]
35. Prochazka M, Lillioja S, Tait JF, Knowler WC, Mott DM, Spraul M, Bennett PH, Bogardus C. Linkage of chromosomal markers on 4q with a putative gene determining maximal insulin action in Pima Indians. Diabetes. 1993;42:514-519. [Abstract]
36. Mitchell BD, Kammerer CM, O'Connell P, Harrison CR, Manire M, Shipman P, Moyer M, Stern MP, Frazier ML. Evidence for linkage of post-challenge insulin levels with intestinal fatty acid binding protein (FABP2) in Mexican Americans. Diabetes. 1995;44:1046-1053. [Abstract]
37. Sellers TA, Drinkard C, Rich SS, Potter JD, Jeffery RW, Hong C-P, Folsom AR. Familial aggregation and heritability of waist-to-hip ratio in adult women: the Iowa Women's Health Study. Int J Obes. 1994;18:607-613.
38. Comuzzie AG, Towne B, Blangero J, Mahaney MC, Chumlea WC, Roche AF, Siervogel RM, MacCluer JW. Differences between Mexican American and non-Hispanic whites in the genetic and environmental correlations between waist and hip circumference: implications for cross-population comparisons of the waist/hip ratio. Am J Hum Biol. 1995;7:120. Abstract.
39. Hollenbeck CB, Chen N, Chen YDI, Reaven GM. Relationship between plasma insulin response to oral glucose and insulin-stimulated glucose utilization in normal subjects. Diabetes. 1984;33:460-463. [Abstract]
40. Mahler RJ. The pathogenesis of pancreatic islet cell hyperplasia and insulin insensitivity in obesity. In: Levine R, Luft R, eds. Advances in Metabolic Disorders. Vol 7. New York, NY: Academic Press; 1974:213-241.
41. Gardner LI, Stern MP, Haffner SM, Gaskill SP, Hazuda HP, Relethford JH, Eifler CW. Prevalence of diabetes in Mexican Americans: relationship to percent of gene pool derived from native American sources. Diabetes. 1984;33:86-92. [Abstract]
42. Chakraborty R, Ferrell RE, Stern MP, Haffner SM, Hazuda HP, Rosenthal M. Relationship of prevalence of non-insulin dependent diabetes mellitus to Amerindian admixture in the Mexican Americans of San Antonio, Texas. Genet Epidemiol. 1986;3:435-454.[Medline] [Order article via Infotrieve]
This article has been cited by other articles:
 |
|
 |
|
  J. Steinberger, S. R. Daniels, R. H. Eckel, L. Hayman, R. H. Lustig, B. McCrindle, and M. L. Mietus-Snyder Progress and Challenges in Metabolic Syndrome in Children and Adolescents: A Scientific Statement From the American Heart Association Atherosclerosis, Hypertension, and Obesity in the Young Committee of the Council on Cardiovascular Disease in the Young; Council on Cardiovascular Nursing; and Council on Nutrition, Physical Activity, and Metabolism Circulation, February 3, 2009; 119(4): 628 - 647. [Full Text] [PDF]
|
|
|
|
 |
|
 P.-I Lin and B. D. Mitchell Approaches for Unraveling the Joint Genetic Determinants of Schizophrenia and Bipolar Disorder Schizophr Bull, July 1, 2008; 34(4): 791 - 797. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Gilibert, A. E. Kwitek, N. Hubner, M. Tschannen, H. J. Jacob, J. Sassard, and A. Bataillard Effects of chromosome 17 on features of the metabolic syndrome in the Lyon hypertensive rat Physiol Genomics, April 1, 2008; 33(2): 212 - 217. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Tang, Y. Hong, M. A. Province, S. S. Rich, P. N. Hopkins, D. K. Arnett, J. S. Pankow, M. B. Miller, and J. H. Eckfeldt Familial Clustering for Features of the Metabolic Syndrome: The National Heart, Lung, and Blood Institute (NHLBI) Family Heart Study. Diabetes Care, March 1, 2006; 29(3): 631 - 636. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Nitsch, M. Molokhia, L. Smeeth, B. L. DeStavola, J. C. Whittaker, and D. A. Leon Limits to Causal Inference based on Mendelian Randomization: A Comparison with Randomized Controlled Trials Am. J. Epidemiol., March 1, 2006; 163(5): 397 - 403. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Chen, S. R. Srinivasan, S. Li, J. Xu, and G. S. Berenson Metabolic Syndrome Variables at Low Levels in Childhood Are Beneficially Associated With Adulthood Cardiovascular Risk: The Bogalusa Heart Study Diabetes Care, January 1, 2005; 28(1): 126 - 131. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. C.Y. Ng, W.-Y. So, V. K.L. Lam, C. S. Cockram, G. I. Bell, N. J. Cox, and J. C.N. Chan Genome-wide Scan for Metabolic Syndrome and Related Quantitative Traits in Hong Kong Chinese and Confirmation of a Susceptibility Locus on Chromosome 1q21-q25 Diabetes, October 1, 2004; 53(10): 2676 - 2683. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. E. Sonnenberg, G. R. Krakower, L. J. Martin, M. Olivier, A. E. Kwitek, A. G. Comuzzie, J. Blangero, and A. H. Kissebah Genetic determinants of obesity-related lipid traits J. Lipid Res., April 1, 2004; 45(4): 610 - 615. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. S. Pankow, D. R. Jacobs Jr, J. Steinberger, A. Moran, and A. R. Sinaiko Insulin Resistance and Cardiovascular Disease Risk Factors in Children of Parents With the Insulin Resistance (Metabolic) Syndrome Diabetes Care, March 1, 2004; 27(3): 775 - 780. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. J. F. Loos, P. T. Katzmarzyk, D. C. Rao, T. Rice, A. S. Leon, J. S. Skinner, J. H. Wilmore, T. Rankinen, and C. Bouchard Genome-Wide Linkage Scan for the Metabolic Syndrome in the HERITAGE Family Study J. Clin. Endocrinol. Metab., December 1, 2003; 88(12): 5935 - 5943. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Tang, M. B. Miller, S. S. Rich, K. E. North, J. S. Pankow, I. B. Borecki, R. H. Myers, P. N. Hopkins, M. Leppert, and D. K. Arnett Linkage Analysis of a Composite Factor for the Multiple Metabolic Syndrome: The National Heart, Lung, and Blood Institute Family Heart Study Diabetes, November 1, 2003; 52(11): 2840 - 2847. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Dallongeville, N. Helbecque, D. Cottel, P. Amouyel, and A. Meirhaeghe The Gly16->Arg16 and Gln27->Glu27 Polymorphisms of {beta}2-Adrenergic Receptor Are Associated with Metabolic Syndrome in Men J. Clin. Endocrinol. Metab., October 1, 2003; 88(10): 4862 - 4866. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. L. Hanson, G. Imperatore, P. H. Bennett, and W. C. Knowler Components of the "Metabolic Syndrome" and Incidence of Type 2 Diabetes Diabetes, October 1, 2002; 51(10): 3120 - 3127. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. B. Meigs, C. I. M. Panhuysen, R. H. Myers, P. W.F. Wilson, and L. A. Cupples A Genome-Wide Scan for Loci Linked to Plasma Levels of Glucose and HbA1c in a Community-Based Sample of Caucasian Pedigrees: The Framingham Offspring Study Diabetes, March 1, 2002; 51(3): 833 - 840. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Arya, J. Blangero, K. Williams, L. Almasy, T. D. Dyer, R. J. Leach, P. O'Connell, M. P. Stern, and R. Duggirala Factors of Insulin Resistance Syndrome-Related Phenotypes Are Linked to Genetic Locations on Chromosomes 6 and 7 in Nondiabetic Mexican-Americans Diabetes, March 1, 2002; 51(3): 841 - 847. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. D Mills and P. J Grant Insulin resistance, haemostatic factors and cardiovascular risk The British Journal of Diabetes & Vascular Disease, January 1, 2002; 2(1): 19 - 26. [Abstract] [PDF]
|
|
|
|
 |
|
 A. H. Kissebah, G. E. Sonnenberg, J. Myklebust, M. Goldstein, K. Broman, R. G. James, J. A. Marks, G. R. Krakower, H. J. Jacob, J. Weber, et al. Quantitative trait loci on chromosomes 3 and 17 influence phenotypes of the metabolic syndrome PNAS, December 19, 2000; 97(26): 14478 - 14483. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Imperatore, W. C. Knowler, D. J. Pettitt, S. Kobes, J. H. Fuller, P. H. Bennett, and R. L. Hanson A Locus Influencing Total Serum Cholesterol on Chromosome 19p : Results From an Autosomal Genomic Scan of Serum Lipid Concentrations in Pima Indians Arterioscler. Thromb. Vasc. Biol., December 1, 2000; 20(12): 2651 - 2656. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. E. Hixson, L. Almasy, S. Cole, S. Birnbaum, B. D. Mitchell, M. C. Mahaney, M. P. Stern, J. W. MacCluer, J. Blangero, and A. G. Comuzzie Normal Variation in Leptin Levels Is Associated with Polymorphisms in the Proopiomelanocortin Gene, POMC J. Clin. Endocrinol. Metab., September 1, 1999; 84(9): 3187 - 3191. [Abstract] [Full Text]
|
|
|
|
|
|
Y. Hong, T. Rice, J. Gagnon, J.-P. Després, A. Nadeau, L. Pérusse, C. Bouchard, A. S. Leon, J. S. Skinner, J. H. Wilmore, et al. Familial Clustering of Insulin and Abdominal Visceral Fat: The HERITAGE Family Study J. Clin. Endocrinol. Metab., December 1, 1998; 83(12): 4239 - 4245. [Abstract] [Full Text]
|
|
|
|
|
|
D. L. Rainwater, B. D. Mitchell, M. C. Mahaney, and S. M. Haffner Genetic Relationship Between Measures of HDL Phenotypes and Insulin Concentrations Arterioscler. Thromb. Vasc. Biol., December 1, 1997; 17(12): 3414 - 3419. [Abstract] [Full Text]
|
|
|
|
|
|
Y. Hong, N. L. Pedersen, N. Egberg, and U. de Faire Moderate Genetic Influences on Plasma Levels of Plasminogen Activator Inhibitor-1 and Evidence of Genetic and Environmental Influences Shared by Plasminogen Activator Inhibitor-1, Triglycerides, and Body Mass Index Arterioscler. Thromb. Vasc. Biol., November 1, 1997; 17(11): 2776 - 2782. [Abstract] [Full Text]
|
|
|
|
|
|
L. Perusse, T. Rice, J. P. Despres, D. C. Rao, and C. Bouchard Cross-Trait Familial Resemblance for Body Fat and Blood Lipids: Familial Correlations in the Quebec Family Study Arterioscler. Thromb. Vasc. Biol., November 1, 1997; 17(11): 3270 - 3277. [Abstract] [Full Text]
|
|
|
|
|
|
W. Group, S. S. Gidding, R. L. Leibel, S. Daniels, M. Rosenbaum, L. Van Horn, and G. R. Marx Understanding Obesity in Youth: A Statement for Healthcare Professionals From the Committee on Atherosclerosis and Hypertension in the Young of the Council on Cardiovascular Disease in the Young and the Nutrition Committee, American Heart Association Circulation, December 15, 1996; 94(12): 3383 - 3387. [Full Text]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||