Lack of Association Between Vitamin D Receptor Genotypes and Haplotypes With Fat-Free Mass in Postmenopausal Brazilian Women

Abstract

The relationship between vitamin D receptor (VDR) ApaI, CDX2, BsmI, FokI, and TaqI polymorphisms and fat-free mass (FFM) were
examined in 191 postmenopausal Brazilian women (mean age 67.87 ± 5.22 years). Participants underwent FFM measurements by dual-energy
x-ray absorptiometry (DEXA). Whole-blood-extracted genomic DNA was genotyped to the aforementioned polymorphisms and to ancestry-informative
markers through minisequencing, using the SNaPshot Multiplex System. Association between VDR polymorphisms and FFM variables
was assessed by analysis of covariance. Haplotypes were estimated, and regression-based, haplotype-specific association tests
were carried out with the studied phenotypes. No departure from Hardy–Weinberg equilibrium was detected for any polymorphism.
None of the investigated VDR allelic variations, individually or analyzed as haplotypes, was associated with FFM phenotypes.
The inclusion of individual African genomic ancestry was used as an attempt to correct for population stratification. Further
studies in larger sample population are required to confirm these findings.

Full text

Lack of Association Between Vitamin D Receptor
Genotypes and Haplotypes With Fat-Free Mass in
Postmenopausal Brazilian Women
Ricardo Moreno Lima,
1
Breno Silva de Abreu,
2
Paulo Gentil,
1
Tulio Cesar de Lima Lins,
2
Da´rio Grattapaglia,
2
Rinaldo Wellerson Pereira,
1,2
and Ricardo Jaco´ de Oliveira
1
1
Programa de Po´s-Graduac¸a˜o em Educac¸a˜o Fı´sica da Universidade Cato´lica de Brası´lia, Brazil.
2
Programa de Po´s-Graduac¸a˜o em Cieˆncias Genoˆmicas e Biotecnologia da Universidade Cato´lica de Brası´lia, Brazil.
The relationship between vitamin D receptor (VDR) ApaI, CDX2, BsmI, FokI, and TaqI
polymorphisms and fat-free mass (FFM) were examined in 191 postmenopausal Brazilian
women (mean age 67.87 6 5.22 years). Participants underwent FFM measurements by dual-
energy x-ray absorptiometry (DEXA). Whole-blood-extracted genomic DNA was genotyped to
the aforementioned polymorphisms and to ancestry-informative markers through minisequenc-
ing, using the SNaPshot Multiplex System. Association between VDR polymorphisms and FFM
variables was assessed by analysis of covariance. Haplotypes were estimated, and regression-
based, haplotype-specific association tests were carried out with the studied phenotypes. No
departure from Hardy–Weinberg equilibrium was detected for any polymorphism. None of the
investigated VDR allelic variations, individually or analyzed as haplotypes, was associated with
FFM phenotypes. The inclusion of individual African genomic ancestry was used as an attempt to
correct for population stratification. Further studies in larger sample population are required to
confirm these findings.
S
ARCOPENIA, defined as the age-associated progressive
loss of skeletal muscle mass and strength (1), is a well
documented phenomenon strongly related to physical dis-
ability among elderly persons (2–6). The health care costs
attributable to sarcopenia in the United States during 2000
were estimated to be around $18.5 billion ($10.8 billion in
men, $7.7 billion in women) (7). Therefore, development of
strategies to minimize skeletal muscle decline will improve
quality of life in elderly persons and decrease public health
care costs. In order to reach this aim, detailed comprehen-
sion of sarcopenia mechanisms is mandatory.
Loss of fat-free mass (FFM) associated with sarcopenia is
a typical complex phenotype in which multiple environ-
mental and genetic factors are thought to interact in its path
(8–11). Nutritional habits and physical activity status are
major environmental determinants of FFM loss (12), and
in a broad perspective, both play their roles through our
genome (13). The human genome normal variation in-
fluence on FFM is poorly understood, and a small number
of candidate genes have been investigated so far (14).
Nevertheless, new technologies emerging from the human
postgenomic era will bring new candidate genes to the field
of sarcopenia genetics (15).
The description that vitamin D deficiency contributes to
age-related muscle function decline (16) and the identi-
fication of vitamin D receptor (VDR) in the nucleus of
myocytes (17,18) introduced the VDR gene as a potential
candidate in association studies involving muscle pheno-
types. It has been suggested that vitamin D modulates
calcium (Ca
2þ
) levels into muscle cells and that nuclear
VDR is the receptor that mediates the effects of this
hormone on contractility (19). In addition, VDR knockout
mice exhibit abnormal muscle development and unregulated
expression of myogenic transcription factors (20).
Despite the above evidence, few studies had investigated
VDR gene variation and its association with muscle strength
and/or FFM. A significant association was found between
BsmI genotypes and muscle strength in postmenopausal
(21) and premenopausal (22) women. However, the G/G
genotype was found to be associated with higher muscle
strength in the postmenopausal women, whereas the oppo-
site association was observed in the premenopausal women.
Regarding other frequently investigated VDR polymor-
phisms, the results are also controversial. There are results
showing an absence of association between ApaI, TaqI, and
FokI VDR polymorphisms and muscle strength in young
and older men (23). Conversely, there are results showing
that ApaI was associated with muscle strength in Chinese
women (24) and that the FokI polymorphism was associ-
ated with FFM in older Caucasian men (25). The CDX2
(G to A substitution) is another functional VDR poly-
morphism frequently investigated in bone phenotype var-
iation (26), but it has not been examined in relation to
muscle phenotypes.
The purpose of the present study was to examine the
association between FFM and TaqI, ApaI, BsmI, FokI, and
CDX2 VDR gene polymorphisms in a sample of post-
menopausal Brazilian women. Aware of the high admixture
966
Journal of Gerontology: BIOLOGICAL SCIENCES Copyright 2007 by The Gerontological Society of America
2007, Vol. 62A, No. 9, 966–972

in Brazilians (27–29) and also aware of the potential com-
plications when performing association studies in such
population (30,31), we genotyped 13 ancestry-informative
single nucleotide polymorphisms (SNPs) and took the esti-
mated genetic ancestry values as covariates during the anal-
ysis. In an effort to get rid of potential factors that may have
influenced FFM variation other than the VDR polymor-
phisms, we also included body mass index, calcium supple-
mentation, smoking status, percent body fat, and hormone
replacement therapy as covariates. The results presented
here failed to detect an association between VDR poly-
morphisms and FFM in postmenopausal Brazilian women.
M
ETHODS
Participants
Participants in the present study were recruited from a full
assistance for the elderly program developed at the Catholic
University of Brasilia, which offers physical activity,
psychological and medical assistance, nutritional assess-
ment, and English classes to the local elderly population.
The present cross-sectional study involved 191 postmeno-
pausal women aged between 56 and 84 years. The 191
volunteers did not have metallic implants, artificial pace-
makers, or hip replacement surgery. They also did not have
a metabolic or endocrine disorder known to affect the
musculoskeletal system. Prior to venous blood collection, all
individuals answered a questionnaire addressing medical
history, comorbidities, hormone replacement therapy, life-
style habits, and self-reported skin color. Another question-
naire was applied to assess physical activity level. All
volunteers provided written informed consent approved by
the Institutional Review Board.
Anthropometric Measurements and Body Composition
Standard procedures were used to gauge weight with 0.1-
kg precision on a physician’s balance beam scale, and height
was measured at the nearest 0.1 cm with a stadiometer.
Body mass index (BMI) was derived as body weight divided
by height squared (kg/m
2
).
Body composition measurements were conducted using
dual-energy x-ray absorptiometry (DEXA; DPX-L; Lunar
Radiation Corporation, Madison, WI). All measurements
were carried out by the same expert technician to avoid
interpretation errors. Regional measurements (arms, legs,
and trunk) were determined on the basis of bone landmarks,
with vertical boundaries separating the arms from the body
at the shoulder, and angled boundaries separating the legs
from the trunk at the hips. Appendicular FFM (AFFM) was
calculated as the sum of both arms’ and legs’ FFM (32).
Because absolute FFM correlates directly with height,
whole-body FFM and AFFM were also considered relative
to body height squared (kg/m
2
), analogous to the use of
BMI (33). In addition, a second measure of relative FFM,
fat-adjusted AFFM, was defined with the use of a linear
regression that predicted volunteers’ AFFM [AFFM (kg) ¼
14.15 þ 18.14 3 height þ 0.09 3 fat mass] from height (in
meters) and whole-body fat mass (in kg), as proposed by
Newman and colleagues (34). The residuals of the re-
gression were used in subsequent analyses. This measure
has been shown to be related to functional limitations (34)
and to markers of inflammation (35) in older individuals.
Physical Activity Level
The official Portuguese short version of the International
Physical Activity Questionnaire (IPAQ) was applied to
assess the physical activity level of each postmenopausal
woman volunteer in this study. The questionnaire was
administered at face-to-face interviews, as is recommended
for use in developing countries. The IPAQ was developed
by investigators from all over the world supported by the
World Health Organization as a tool to follow up physical
activity level. It has been reported as an instrument that has
acceptable measurement properties in various countries
(36), and it has been previously applied in a postmenopausal
Brazilian population study (37). Based on the questionnaire
results, all individuals were categorized as sedentary,
insufficiently active, active, or very active.
Genotyping
The polymorphisms analyzed in this study are commonly
reported as TaqI, ApaI, BsmI, FokI, and CDX2. These
polymorphisms are found in the National Center for
Biotechnology Information (NCBI) dbSNP data bank
(http://www.ncbi.nlm.nih.gov/projects/SNP/) under the re-
spective denominations: rs731236, rs7975232, rs1544410,
rs10735810, and rs11568820. In order to make comprehen-
sion easier, the present study uses the restriction fragment
length polymorphism (RFLP; TaqI, ApaI, BsmI, and FokI)
and the transcription factor binding site (CDX2) nomencla-
ture to refer to each one of the polymorphisms. The alleles,
genotypes, and haplotypes are described using the dbSNP
reference alleles.
Genotyping was performed on DNA extracted from pe-
ripheral venous blood using a modified salting-out protocol
(38). The polymerase chain reaction (PCR) protocol was
carried out in 12.5 lL as follows: 1X Taq polymerase
buffer, 2.5 mM MgCl
2
, 250 mM dNTPs, bovine serum
albumin (BSA) at 1.6 mg/mL, 0.50 lM of each primer,
10–40 ng of DNA and 1 U of Taq polymerase. The PCR
amplification was performed in an ABI9700 thermocycler
using an initial denaturation step at 958C for 5 minutes,
followed by 15 cycles of 40 seconds at 958C, 40 seconds
at 628C decreasing 0.58C per cycle, 40 seconds at 728C, and
15 cycles of 40 seconds at 958C, 40 seconds at 568C,
40 seconds at 728C, and a final extension step at 728C for
5 minutes. The purification was carried out on 3 lL of PCR
volume adding 1 U of ExoI, 0.95 U of SAP, and 0.5 X SAP
Reaction Buffer, which were incubated for 90 minutes at
378C following 20 minutes at 808C. The minisequencing
was performed using 1.25 lL of SNaPshot Multiplex
minisequencing kit reaction mix (Applied Biosystems,
Foster City, CA), 1.25 lL of Big Dye Sequencing Buffer,
1 lL of Purified PCR product, 0.4 lM single base extension
primers each, and sterile autoclaved Milli-Q water up to 5
lL. Reaction conditions were performed as follows: 25
cycles of 968C for 10 seconds, 508C for 5 seconds, and 608C
for 30 seconds. Purification was carried out in order to de-
grade ddNTPs not incorporated on reaction by adding 0.5 U
967FFM AND VDR POLYMORPHISM S IN POSTMENOPAUSAL WOMEN

of SAP and 0.2X SAP reaction buffer for each sample,
followed by an incubation at 378C for 60 minutes and a step
at 758C for 20 minutes. Samples were subject to electro-
phoresis on an ABI prism 3100 Genetic Analyzer (Applied
Biosystems) and analyzed with GeneScan Analysis 3.7 and
Genotyper 3.7 software (Applied Biosystems). Genotyping
of 13 ancestry-informative markers (AIMs) was performed
under the same conditions described above.
Genomic Ancestry Inference
Thirteen AIMs were genotyped for all the samples.
Detailed information about these AIMs are provided in
Table 1. The up-to-date frequencies of the AIMs on all
parental populations used in the computational inference of
genomic ancestry were retrieved from the NCBI dbSNP,
and such analysis was performed by the LOD-Score–based
software IAE3CI, which was provided by Dr. Mark D.
Shriver (Pennsylvania State University). This analysis
software estimates the amount of genomic contribution
from each of the parental population on one’s genome.
Statistical Analysis
The Kolmogorov–Smirnov test was used to verify data
distribution normality. Compliance of VDR genotype
frequencies to Hardy–Weinberg equilibrium expectancy
were analyzed by exact test. Each polymorphism was
individually analyzed using one-way analysis of variance
(ANOVA) models to test for differences in physical activity
levels, age, height, BMI, genomic ancestry levels, and
percent body fat among VDR genotype groups. A stepwise
multiple regression model including all potential covariates
was analyzed for each of the investigated FFM variables
when searching for differences between VDR genotype
groups in AFFM, relative AFFM, whole-body FFM, relative
whole-body FFM, and fat-adjusted AFFM. Variables with
p , .10 were included as covariates in the subsequent anal-
ysis of covariance (ANCOVA) performed for each poly-
morphism separately. The Bonferroni procedure was
adopted to correct for multiple comparisons. Data were
considered significant at p , .05. All the above statistical
analyses were performed using the Statistical Package for
the Social Sciences, version 10.0 (SPSS, Chicago, IL).
All haplotype estimates and regression-based, haplotype-
specific association tests were carried out using Whap
(http://www.genome.wi.mit.edu/;shaun/whap/) software,
which is based on an expectation-maximization (EM)
algorithm and considers the estimates of posterior probabil-
ities to account for the ambiguity of haplotype phase
estimates on regression-based association tests on unrelated
individuals. The software package is able to handle
quantitative traits and covariates for regression analysis.
The haplotype association analyses were performed consid-
ering whole-body FFM, AFFM, relative whole-body FFM,
relative AFFM, and fat-adjusted AFFM as dependent
variables. The model used included height, weight, percent
body fat, and individual African ancestry levels.
R
ESULTS
Baseline Characteristics
The Kolmogorov–Smirnov test showed that all variables
were normally distributed. Therefore, all genotyped partic-
ipants were included in subsequent analyses. The popula-
tion’s physical characteristics are outlined in Table 2. It was
observed that 55 (21.2%) participants were classified as
Table 1. Thirteen Ancestry-Informative Markers with Chromosomal Positions, Allele Frequencies, and Their Respective Differences (d)
Between European (EUR), African (AFR), and Amerindians (AMR)
Allele Frequency Allele Frequency Differences (d)
Locus Position Allele EUR AFR AMR EUR/AFR EUR/AMR AFR/AMR
rs3138521 1q25 Insertion 0.280 0.860 0.061 0.580 0.219 0.799
rs3176921 8q13.1 G 0.073 0.680 0.017 0.607 0.056 0.663
rs2740574 7q22.1 G 0.042 0.802 0.041 0.760 0.001 0.761
rs2814778 1q23.2 C 0.998 0.001 0.998 0.997 0.000 0.997
rs285 8p21.3 G 0.508 0.029 0.558 0.479 0.050 0.529
rs1800404 15q13.1 G 0.210 0.910 0.570 0.700 0.360 0.340
rs1129048 1p35 C 0.224 0.995 0.983 0.771 0.759 0.012
rs1426654 15q21 C 0.001 0.980 0.950 0.979 0.949 0.030
rs1480642 6q23 C 0.994 0.106 0.621 0.888 0.373 0.515
rs1871534 5p15.2 C 0.999 0.017 0.999 0.982 0.000 0.982
rs3768641 2p13 C 0.083 0.999 0.068 0.916 0.015 0.931
rs4766807 12q24.2 T 0.378 0.970 0.052 0.592 0.326 0.918
rs735665 11q24 C 0.244 0.939 0.018 0.695 0.226 0.921
Table 2. Volunteers’ Characteristics
Variable
N 191
Age, y 67.87 6 5.22
Weight, kg 64.55 6 11.73
Height, m 1.52 6 0.06
BMI, kg/m
2
27.91 6 4.49
AFFM, kg 15.85 6 2.65
Relative AFFM, kg/m
2
6.85 6 0.95
Whole-body FFM, kg 38.20 6 5.23
Relative whole-body FFM, kg/m
2
16.52 6 1.82
Percent body fat, % 38.29 6 6.35
Notes: Values are expressed as the mean 6 standard deviation.
AFFM ¼ appendicular fat-free mass; FFM ¼ fat-free mass; BMI ¼ body
mass index.
968 LIMA ET AL.

obese (BMI . 30 kg/m
2
), and 33 (17.28%) were taking
hormone replacement therapy. The most common chronic
diseases verified among the volunteers were hypertension
and type II diabetes, respectively, affecting 127 (66.5%) and
32 (16.8%) volunteers. Physical activity levels according to
IPAQ classification were as follows: 8 (4.2%) were sed-
entary, 67 (35.1%) were insufficiently active, 113 (59.1%)
were active, and 3 (1.6%) were very active.
Stepwise multivariate regression analysis revealed that
height, weight, and body fat percentage were the most
important predictors of AFFM and whole-body FFM,
whereas the model that better predicted relative AFFM
and relative whole-body FFM included BMI and percent
body fat. For fat-adjusted AFFM, the significant predictor
was BMI. Consequently, these variables were used as
covariates in subsequent analyses of covariance, along with
the individual African ancestry level.
Genotyping
VDR genotype results were available for the majority of
participants. However, in some samples it could not be
precisely identified, so those were not included in
association analyses. Reasons for incomplete genotype data
included unsuccessful PCR assays or single base extension
reaction. Therefore, when presenting the results of each
locus, the numbers of participants are slightly different.
More precisely, a total of 187, 183, 189, 189, and 184
postmenopausal women were genotyped for the VDR ApaI,
BsmI, CDX2, FokI, and TaqI polymorphisms, respectively.
Allelic frequencies and genotypic distribution are presented
in Tables 3 and 4, respectively. No departures from Hardy–
Weinberg equilibrium were detected for any of the studied
polymorphisms.
Baseline Characteristics and FFM Phenotypes in
Relation to VDR Genotype
The studied population characteristics according to VDR
genotypes are presented in Table 4. No significant differ-
ences were observed among VDR polymorphism genotype
distribution regarding age, height, weight, BMI, ancestry
levels, or percent body fat. Additionally, there were no
significant differences between VDR genotypes for physical
activity level or hormone replacement therapy.
As shown in Table 5, ANCOVA revealed no statistically
significant difference between ApaI, BsmI, CDX2, FokI, or
TaqI VDR genotypes for AFFM, relative AFFM, whole-body
FFM, relative whole-body FFM, or fat-adjusted AFFM.
Haplotype Association Tests
Haplotype estimates from TaqI, ApaI, BsmI, FokI, and
CDX2 polymorphism genotypes revealed 15 haplotypes
with frequencies . 2% spanning 96.4% of haplotypic
diversity (Table 6). Omnibus permutation tests for haplo-
typic association were carried out considering both the five
marker haplotypes and the 39 end TaqI-ApaI-BsmI hap-
lotypes. The regression-based association tests for the TaqI,
ApaI, and BsmI haplotype did not give significant results for
any of the phenotypic traits described here (p ¼ 1).
No significant results from the association analyses of
AFFM, FFM, relative AFFM, relative FFM, and fat-adjusted
AFFM to the five marker haplotypes were detected when no
covariates were included in the model (p ¼ 1). The inclusion
of each of the covariates alone to correct for height, weight,
BMI, and percent body fat and all possible combinations
among them were not able to modify the previous observation
(p ¼ 1). The inclusion of Africanicity level as a measure to
correct for population stratification was also unable to
evidence haplotype association as well (p ¼ 1).
D
ISCUSSION
The role played by the vitamin D endocrine system in
muscle mass is clearly shown in VDR knockout mice (20).
Although it is known that vitamin D shows its genomic and
rapid response effects through nuclear and membrane-bound
VDR, little is known about the role played by VDR gene
normal variation over muscle phenotypes, more specifically,
the loss of muscle mass and strength associated with aging.
So far, this question has been addressed through genetic
association studies investigating the VDR polymorphisms
ApaI, TaqI, BsmI, FokI, 39 poly A repeat, and their
association with muscle mass and strength (21–25). By the
time this article was submitted, we were not aware of any
publication that evaluated the association between the
CDX2 polymorphism and muscle phenotypes. In genotyp-
ing ApaI, TaqI, BsmI, FokI, and CDX2 polymorphisms in
a sample of postmenopausal Brazilian women, we did not
find an association with FFM phenotypes. The association
was not found with any allele, genotype, or haplotype.
In fact, available data do not support evidence of asso-
ciations among TaqI genotypes and muscular phenotypes
(24,39). We are not aware of association studies investigating
ApaI polymorphism and FFM. Its relationship with muscle
strength has been previously investigated; however, the
results are not consistent. For example, Iki and colleagues
(39) observed no differences in any of various analyzed
muscle strength indices among ApaI genotype groups in a
sample of 180 postmenopausal Japanese women. Because
muscle strength is positively related with FFM (40), these
results may support our observation. In contrast, a recent
investigation (24) reported differences among ApaI genotype
groups for isokinetic muscle strength. However, the authors
Table 3. Allelic Frequencies and dbSNP Numbers of the Studied
Polymorphisms in the VDR Gene
SNP dbSNP Allele Frequency
ApaI (A/C) rs7975232 A 0.586
C 0.414
BsmI (A/G) rs1544410 A 0.317
G 0.683
Cdx-2 (G/A) rs11568820 G 0.646
A 0.354
FokI (C/T) rs10735810 C 0.648
T 0.352
TaqI (T/C) rs731236 T 0.668
C 0.332
Note: dbSNP ¼ single nucleotide polymorphism data bank (National Center
for Biotechnology Information); VDR ¼ vitamin D receptor.
969FFM AND VDR POLYMORPHISM S IN POSTMENOPAUSAL WOMEN

compared the A/A homozygous group to the C/C and A/C
groups combined, in a different statistical approach. In
agreement with our results, Roth and colleagues (25) reported
that the BsmI polymorphism was not significantly associated
with absolute or relative FFM in a cohort of older Caucasian
men. Similar findings were observed by Grundberg and
colleagues (22), who reported no statistical differences
among VDR BsmI genotype groups for whole-body FFM
in a study sample of 175 Swedish women aged 20–39 years.
Contrary to our findings, Roth and colleagues (25) reported
significant FokI genotype differences for each of the same
FFM variables evaluated in the present study. In elderly
Caucasian men, the authors observed that the C/C group
exhibited significantly lower values than both the C/T and T/T
groups for all the examined FFM variables. To our
knowledge, no other study examined the relationship between
FFM and the CDX2 polymorphism.
Nonreplication is the rule, not the exception, when
dealing with association studies’ results. This is currently
explained by aspects of the study design itself, by the
genetic architecture of the trait under investigation, and most
frequently by the combination of both (41). Population
heterogeneity may play an important role in association
studies contributing to spurious results, not only false-
positives but also false-negatives. Aware of the admixed
nature of the Brazilian population, we used African genomic
ancestry as a covariate in an attempt to address this potential
source of bias. Such a procedure intended to minimize the
well characterized problem generated from the use of self-
reports of ancestry or skin pigmentation on sample clus-
tering and analysis in association studies (42).
The results presented here, those already published, and
others that may appear in the literature should, in the future,
be summed up and submitted to meta-analysis as already
happened in the case of VDR variation and bone phenotypes
(43–46). Nevertheless, meta-analyses are prone to biases
resulting from selective publication of positive results;
therefore, publications of negative associations are needed
to be considered by the scientific community. By now, the
number of published articles investigating VDR gene
variation and muscle phenotypes is much lower than those
in which bone phenotypes were investigated. Therefore, it is
Table 4. Characteristics of the Participants According to ApaI, CDX2, BsmI, FokI, and TaqI Vitamin D Receptor Genotypes
ApaI genotype C/C A/C A/A p Value
N, % 29 (15.51) 97 (51.87) 61 (32.62)
Age, y 67.86 6 5.37 67.69 6 5.12 68.21 6 5.34 .829
Weight, kg 62.69 6 11.49 65.54 6 11.68 63.85 6 12.21 .450
Height, cm 151.93 6 5.99 152.15 6 6.26 151.65 6 6.75 .878
BMI, kg/m
2
27.09 6 4.17 28.32 6 4.86 27.63 6 3.99 .374
Percent body fat, % 38.64 6 6.25 38.16 6 6.31 38.47 6 6.41 .882
BsmI genotype G/G G/A A/A
N, % 90 (49.18) 70 (38.25) 23 (12.57)
Age, y 68.02 6 5.27 67.50 6 5.33 67.13 6 4.30 .695
Weight, kg 64.59 6 11.42 65.74 6 13.00 62.13 6 10.08 .447
Height, cm 151.43 6 6.09 153.03 6 6.89 151.00 6 5.33 .210
BMI, kg/m
2
28.18 6 4.76 27.94 6 4.47 27.17 6 3.55 .635
Percent body fat, % 38.71 6 6.12 38.36 6 6.24 37.31 6 6.22 .623
CDX2 genotype Cdx-G Cdx G/A Cdx-A
N, % 79 (41.80) 86 (45.50) 24 (12.70)
Age, y 67.65 6 4.89 67.91 6 5.86 68.38 6 3.94 .832
Weight, kg 63.56 6 11.21 66.06 6 11.72 62.88 6 13.53 .294
Height, cm 151.66 6 7.00 152.48 6 5.53 150.63 6 6.85 .410
BMI, kg/m
2
27.58 6 4.23 28.39 6 4.57 27.60 6 5.07 .474
Percent body fat, % 38.56 6 5.88 38.32 6 6.11 37.96 6 8.26 .915
FokI genotype C/C C/T T/T
N, % 79 (41.80) 87 (46.03) 23 (12.17)
Age, y 67.94 6 5.33 67.68 6 5.02 68.48 6 5.95 .805
Weight, kg 63.77 6 10.94 65.25 6 12.01 64.65 6 13.81 .721
Height, cm 152.38 6 5.54 151.37 6 6.72 152.00 6 7.29 .556
BMI, kg/m
2
27.43 6 4.27 28.41 6 4.51 27.94 6 5.17 .380
Percent body fat, % 37.94 6 6.62 39.06 6 5.99 37.20 6 6.23 .332
TaqI genotype C/C C/T T/T
N, % 28 (15.22) 66 (35.87) 90 (48.91)
Age, y 67.54 6 5.66 67.44 6 5.14 68.32 6 5.17 .542
Weight, kg 62.89 6 10.04 65.03 6 11.26 64.52 6 12.42 .718
Height, cm 151.25 6 6.26 153.29 6 6.21 151.28 6 6.44 .120
BMI, kg/m
2
27.38 6 3.17 27.58 6 3.85 28.19 6 5.12 .582
Percent body fat, % 38.75 6 5.49 37.36 6 6.02 38.67 6 6.67 .388
Note: BMI ¼ body mass index.
970 LIMA ET AL.

necessary that more research groups investigate and publish
their results, so that a clearer picture of the relationship
between VDR gene natural variation and muscle mass loss
phenotypes can be seen.
Summary
The present results indicate no association between the
ApaI, CDX2, BsmI, FokI, and TaqI polymorphisms in the
VDR gene, individually or when analyzed as haplotypes,
with indices of FFM in postmenopausal Brazilian women.
These results remained unchanged with the inclusion of
Africanicity levels as a measure to correct for population
stratification. Further studies in larger sample populations
are required to confirm these findings.
ACKNOWLEDGMENTS
This work was supported by Coordenac¸a˜o de Aperfeic¸oamento de
Pessoal de Nı´vel Superior (CAPES), Conselho Nacional de Desenvolvi-
mento Cientı´fico e Tecnolo´gico (CNPq), and Universidade Cato´lica de
Brası´lia (PRPGP–UCB). TCLL and BSA were supported by a CAPES
scholarship.
We thank Dr. Mark D. Shriver for providing IA3CI software and
Dr. Shaun Purcell for helping with the Whap software.
C
ORRESPONDENCE
Address correspondence to Ricardo Jaco´ de Oliveira, PhD, Universidade
Cato´lica de Brası´lia–UCB, Mestrado em Educac¸a˜o Fı´sica, QS 07, Lote 01,
Pre´dio Sa˜o Joa˜o Bosco, Sala 119. CEP: 71.996-700, Taguatinga–DF–
Brazil. E-mail: rjaco@ucb.br or rjaco@zaz.com.br
Table 5. Relationship Between Fat-Free Mass Variables (Mean 6 Standard Deviation [SD]) and ApaI, Cdx-2, BsmI, FokI,
or TaqI VDR Genotype Groups in the Study Population, by Analysis of Covariance
VDR Gene ApaI Genotype Groups C/C A/C A/A p Value
AFFM, kg 15.37 6 2.48 16.03 6 2.92 15.78 6 2.35 .829
Relative AFFM, kg/m
2
6.64 6 0.85 6.92 6 1.10 6.84 6 0.71 .778
Whole Body FFM, kg 36.91 6 5.33 38.65 6 5.15 37.95 6 5.37 .412
Relative Whole Body FFM, kg/m
2
15.96 6 1.86 16.69 6 1.93 16.46 6 1.59 .285
Fat-adjusted AFFM, kg .33 6 1.82 .02 6 1.56 .09 6 1.58 .586
VDR Gene BsmI Genotype Groups G/G G/A A/A
AFFM, kg 15.67 6 2.04 16.20 6 3.53 15.71 6 1.89 .707
Relative AFFM, kg/m
2
6.81 6 0.76 6.88 6 1.22 6.89 6 0.76 .607
Whole Body FFM, kg 37.78 6 5.07 38.93 6 5.95 37.71 6 3.79 .418
Relative Whole Body FFM, kg/m
2
16.48 6 2.00 16.56 6 1.79 16.52 6 1.22 .322
Fat-adjusted AFFM, kg .05 6 1.54 .06 6 1.69 .27 6 1.60 .484
VDR Gene Cdx-2 Genotype Groups Cdx-G Cdx-G/A Cdx-A
AFFM, kg 15.63 6 2.33 16.17 6 3.06 15.38 6 2.01 .914
Relative AFFM, kg/m
2
6.78 6 0.78 6.94 6 1.14 6.76 6 0.65 .911
Whole Body FFM, kg 37.50 6 5.41 39.12 6 5.17 37.05 6 4.61 .523
Relative Whole Body FFM, kg/m
2
16.28 6 1.85 16.81 6 1.87 16.31 6 1.53 .660
Fat-adjusted AFFM, kg .05 6 1.64 .05 6 1.62 .10 6 1.38 .977
VDR Gene FokI Genotype Groups C/C C/T T/T
AFFM, kg 15.99 6 2.97 15.76 6 2.44 15.67 6 2.40 .303
Relative AFFM, kg/m
2
6.87 6 1.13 6.86 6 0.80 6.77 6 0.83 .287
Whole Body FFM, kg 38.02 6 4.36 38.11 6 5.77 38.92 6 6.05 .957
Relative Whole Body FFM, kg/m
2
16.36 6 1.56 16.60 6 1.98 16.83 6 2.14 .943
Fat-adjusted AFFM, kg .03 6 1.44 .04 6 1.70 .04 6 1.78 .982
VDR Gene TaqI Genotype Groups C/C C/T T/T
AFFM, kg 15.41 6 1.89 16.08 6 2.37 15.85 6 3.06 .649
Relative AFFM, kg/m
2
6.72 6 0.59 6.82 6 0.70 6.92 6 1.18 .746
Whole Body FFM, kg 37.62 6 4.07 39.07 6 5.37 37.71 6 5.32 .108
Relative Whole Body FFM, kg/m
2
16.43 6 1.34 16.58 6 1.59 16.48 6 2.07 .640
Fat-adjusted AFFM, kg .18 6 1.39 .13 6 1.60 .02 6 1.78 .618
Note: AFFM ¼ appendicular fat-free mass; FFM ¼ fat-free mass; VDR ¼ vitamin D receptor.
Table 6. List of Phased Haplotypes Reconstructed Through
Expectation–Maximization Algorithm and
Its Respective Frequencies
ID Haplotype Frequency
1 CAACG 0.166
2 TCGCG 0.153
3 TCGCA 0.121
4 TCGTG 0.096
5 TAGCG 0.066
6 CAATG 0.065
7 TAGTG 0.056
8 CAATA 0.048
9 TAGCA 0.046
10 TCGTA 0.036
11 TAGTA 0.034
12 CAGCA 0.034
13 TAACA 0.033
14 TAACG 0.022
15 CAGTG 0.021
Note: Haplotype follows the gene sequence of markers as TaqI, ApaI, BsmI,
FokI, and CDX2.
971FFM AND VDR POLYMORPHISM S IN POSTMENOPAUSAL WOMEN

REFERENCES
1. Rosenberg IH. Summarycomments.Am J Clin Nutr. 1989;50:1231–1233.
2. Madsen OR, Lauridsen UB, Hartkopp A, Sorensen OH. Muscle
strength and soft tissue composition as measured by dual energy x-ray
absorptiometry in women aged 18–87 years. Eur J Appl Physiol Occup
Physiol. 1997;75:239–245.
3. Baumgartner RN, Waters DL, Gallagher D, Morley JE, Garry PJ.
Predictors of skeletal muscle mass in elderly men and women. Mech
Ageing Dev. 1999;107:123–136.
4. Frontera WR, Suh D, Krivickas LS, Hughes VA, Goldstein R,
Roubenoff R. Skeletal muscle fiber quality in older men and women.
Am J Physiol Cell Physiol. 2000;279:C611–618.
5. Melton LJ 3rd, Khosla S, Crowson CS, O’Connor MK, O’Fallon WM,
Riggs BL. Epidemiology of sarcopenia. J Am Geriatr Soc. 2000;48:
625–630.
6. Iannuzzi-Sucich M, Prestwood KM, Kenny AM. Prevalence of
sarcopenia and predictors of skeletal muscle mass in healthy, older
men and women. J Gerontol Med Sci. 2002;57A:M772–M777.
7. Janssen I, Shepard DS, Katzmarzyk PT, Roubenoff R. The healthcare
costs of sarcopenia in the United States. J Am Geriatr Soc. 2004;52:80–85.
8. Arden NK, Spector TD. Genetic influences on muscle strength, lean
body mass, and bone mineral density: a twin study. J Bone Miner Res.
1997;12:2076–2081.
9. Nguyen TV, Howard GM, Kelly PJ, Eisman JA. Bone mass, lean mass,
and fat mass: same genes or same environments? Am J Epidemiol.
1998;147:3–16.
10. Roubenoff R, Hughes VA. Sarcopenia: current concepts. J Gerontol
Med Sci. 2000;55A:M716–M724.
11. Tiainen K, Sipila S, Alen M, et al. Heritability of maximal isometric
muscle strength in older female twins. J Appl Physiol. 2004;96:173–180.
12. Carmeli E, Coleman R, Reznick AZ. The biochemistry of aging
muscle. Exp Gerontol. 2002;37:477.
13. Booth FW, Chakravarthy MV, Spangenburg EE. Exercise and gene
expression: physiological regulation of the human genome through
physical activity. J Physiol. 2002;543:399–411.
14. Gordon ES, Gordish Dressman HA, Hoffman EP. The genetics of muscle
atrophy and growth: the impact and implications of polymorphisms in
animals and humans. Int J Biochem Cell Biol. 2005;37:2064–2074.
15. Giresi PG, Stevenson EJ, Theilhaber J, et al. Identification of a molecular
signature of sarcopenia. Physiol Genomics. 2005;21:253–263.
16. Bischoff HA, Stahelin HB, Urscheler N, et al. Muscle strength in the
elderly: its relation to vitamin D metabolites. Arch Phys Med Rehabil.
1999;80:54–58.
17. Bischoff HA, Borchers M, Gudat F, et al. In situ detection of 1,25-
dihydroxyvitamin D3 receptor in human skeletal muscle tissue.
Histochem J. 2001;33:19–24.
18. Bischoff-Ferrari HA, Borchers M, Gudat F, Durmuller U, Stahelin HB,
Dick W. Vitamin D receptor expression in human muscle tissue
decreases with age. J Bone Miner Res. 2004;19:265–269.
19. Capiati D, Benassati S, Boland RL. 1,25(OH)2-vitamin D3 induces
translocation of the vitamin D receptor (VDR) to the plasma membrane
in skeletal muscle cells. J Cell Biochem. 2002;86:128–135.
20. Endo I, Inoue D, Mitsui T, et al. Deletion of vitamin D receptor gene in
mice results in abnormal skeletal muscle development with deregulated
expression of myoregulatory transcription factors. Endocrinology.
2003;144:5138–5144.
21. Geusens P, Vandevyver C, Vanhoof J, Cassiman JJ, Boonen S, Raus J.
Quadriceps and grip strength are related to vitamin D receptor genotype
in elderly nonobese women. J Bone Miner Res. 1997;12:2082–2088.
22. Grundberg E, Brandstrom H, Ribom EL, Ljunggren O, Mallmin H,
Kindmark A. Genetic variation in the human vitamin D receptor is
associated with muscle strength, fat mass and body weight in Swedish
women. Eur J Endocrinol. 2004;150:323–328.
23. Van Pottelbergh I, Goemaere S, De Bacquer D, De Paepe A, Kaufman
M. Vitamin D receptor gene allelic variants, bone density, and bone
turnover in community-dwelling men. Bone. 2002;31:631–637.
24. Wang P, Ma LH, Wang HY, et al. Association between polymorphisms
of vitamin D receptor gene ApaI, BsmI and TaqI and muscular strength
in young Chinese women. Int J Sports Med. 2006;27:182–186.
25. Roth SM, Zmuda JM, Cauley JA, Shea PR, Ferrell RE. Vitamin D
receptor genotype is associated with fat-free mass and sarcopenia in
elderly men. J Gerontol A Biol Sci Med Sci. 2004;59:10–15.
26. Arai H, Miyamoto KI, Yoshida M, et al. The polymorphism in the
caudal-related homeodomain protein Cdx-2 binding element in the
human vitamin D receptor gene. J Bone Miner Res. 2001;16:1256–1264.
27. Callegari-Jacques SM, Grattapaglia D, Salzano FM, et al. Historical
genetics: spatiotemporal analysis of the formation of the Brazilian
population. Am J Hum Biol. 2003;15:824–834.
28. Marrero AR, Das Neves Leite FP, De Almeida Carvalho B, et al.
Heterogeneity of the genome ancestry of individuals classified as
White in the state of Rio Grande do Sul, Brazil. Am J Hum Biol. 2005;
17:496–506.
29. Parra FC, Amado RC, Lambertucci JR, Rocha J, Antunes CM, Pena
SD. Color and genomic ancestry in Brazilians. Proc Natl Acad Sci
USA.2003;100:177–182.
30. Choudhry S, Coyle NE, Tang H, et al. Population stratification
confounds genetic association studies among Latinos. Hum Genet.
2006;118:652–664.
31. Hoggart CJ, Parra EJ, Shriver MD, et al. Control of confounding of
genetic associations in stratified populations. Am J Hum Genet. 2003;
72:1492–1504.
32. Heymsfield SB, Smith R, Aulet M, et al. Appendicular skeletal muscle
mass: measurement by dual-photon absorptiometry. Am J Clin Nutr.
1990;52:214–218.
33. Baumgartner RN, Koehler KM, Gallagher D, et al. Epidemiology of
sarcopenia among the elderly in New Mexico. Am J Epidemiol. 1998;
147:755–763.
34. Newman AB, Kupelian V, Visser M, et al. Sarcopenia: alternative
definitions and associations with lower extremity function. JAm
Geriatr Soc. 2003;51:1602–1609.
35. Cesari M, Kritchevsky SB, Baumgartner RN, et al. Sarcopenia, obesity,
and inflammation–results from the Trial of Angiotensin Converting
Enzyme Inhibition and Novel Cardiovascular Risk Factors study. Am J
Clin Nutr. 2005;82:428–434.
36. Craig CL, Marshall AL, Sjostrom M, et al. International physical
activity questionnaire: 12-country reliability and validity. Med Sci
Sports Exerc. 2003;35:1381–1395.
37. da Silva RB, Costa-Paiva L, Pinto-Neto AM, Braga Ade A, Morais SS.
Association between habitual physical activity and parameters of phys-
ical fitness in postmenopausal women. Climacteric. 2005;8:360–370.
38. Miller SA, Dykes DD, Polesky HF. A simple salting out procedure for
extracting DNA from human nucleated cells. Nucleic Acids Res.
1988;16:1215.
39. Iki M, Saito Y, Dohi Y, et al. Greater trunk muscle torque reduces
postmenopausal bone loss at the spine independently of age, body size,
and vitamin D receptor genotype in Japanese women. Calcif Tissue Int.
2002;71:300–307.
40. Lindle RS, Metter EJ, Lynch NA, et al. Age and gender comparisons of
muscle strength in 654 women and men aged 20–93 yr. J Appl Physiol.
1997;83:1581–1587.
41. Ioannidis JP, Trikalinos TA, Khoury MJ. Implications of small effect
sizes of individual genetic variants on the design and interpretation of
genetic association studies of complex diseases. Am J Epidemiol.
2006;164:609–614.
42. Barnholtz-Sloan JS, Chakraborty R, Sellers TA, Schwartz AG.
Examining population stratification via individual ancestry estimates
versus self-reported race. Cancer Epidemiol Biomarkers Prev. 2005;
14:1545–1551.
43. Cooper GS, Umbach DM. Are vitamin D receptor polymorphisms
associated with bone mineral density? A meta-analysis. J Bone Miner
Res. 1996;11:1841–1849.
44. Thakkinstian A, D’Este C, Attia J. Haplotype analysis of VDR gene
polymorphisms: a meta-analysis. Osteoporos Int. 2004;15:729–734.
45. Thakkinstian A, D’Este C, Eisman J, Nguyen T, Attia J. Meta-analysis of
molecular association studies: vitamin D receptor gene polymorphisms
and BMD as a case study. J Bone Miner Res. 2004;19:419–428.
46. Uitterlinden AG, Ralston SH, Brandi ML, et al. The association
between common vitamin D receptor gene variations and osteoporosis:
a participant-level meta-analysis. Ann Intern Med.
2006;145:255–264.
Received November 9, 2006
Accepted June 12, 2007
Decision Editor: Huber R. Warner, PhD
972 LIMA ET AL.