Heritability of Cardiovascular Risk Factors in a Chinese Population – Taichung
Community Health Study and Family Cohort
Cheng-Chieh Lin, 1,2,3 Patricia A. Peyser, 4§ Sharon L.R. Kardia, 4,5 § , Chia-Ing Li, 2,3
Chiu-Shong Liu, 1,2,3 Julia S. Chu,4 Wen-Yuan Lin,1,2,5 Tsai-Chung Li, * 6,7
1. Department of Family Medicine, China Medical University Hospital, Taichung,
Taiwan
2. School of Medicine, College of Medicine, China Medical University, Taichung,
Taiwan
3. Department of Medical Research, China Medical University Hospital, Taichung,
Taiwan
4. Department of Epidemiology, University of Michigan, Ann Arbor, Michigan, USA
5. Graduate Institute of Clinical Medical Science, College of Medicine, China
Medical University, Taichung, Taiwan
6. Graduate Institute of Biostatistics, College of Management, China Medical
University, Taichung, Taiwan
7. Department of Healthcare Administration, College of Health Science, Asia
University, Taichung, Taiwan
Short title: Heritability of cardiovascular risk factors
§: Authors equally contributed to the first author.
Correspondence to: Tsai-Chung Li
China Medical University, 91 Hsueh-Shih Road, Taichung, 40421, Taiwan, Tel:
886-4-2205-3366 ext. 6605, Fax: 886-4-2207-8539, e-mail: tcli@mail.cmu.edu.tw
1
Abstract
Background and aim: Previous studies reporting on estimates of heritability of
cardiovascular risk factors in Chinese are limited. This study aims to estimate the
heritability of cardiovascular risk factors in relatives of residents who participated in
the Taichung Community Health Study (TCHS) and Family Cohort (TCHS-FC) while
controlling as many potential confounders as possible.
Methods: A total of 1564 study subjects from 494 families with members aged 12 to
91 years were enrolled from a random sample of participants of TCHS and their
family members (TCHS-FC) from 2009 to 2012. Anthropometric measurement, body
composition, blood pressure, plasma lipids, fasting glucose, insulin, highly sensitive
C-reactive protein (hs-CRP), brachial-ankle pulse wave velocity (baPWV), and the
ankle-brachial index (ABI), as well as a questionnaire interview, were obtained from
each participant.
Results: Cardiovascular risk factors with estimates of heritability greater than 30%
after multivariate adjustment were triglyceride (h2=0.41), HDL-C (h2=0.49), LDL-C
(h2=0.47), total cholesterol (h2=0.46), hip circumference (h2=0.44), weight (h2=0.42),
insulin (h2=0.39), hs-CRP (h2=0.38), BMI (h2=0.38), and percent body fat mass
(h2=0.35). Correlation coefficients for significant sibling varied from 0.10 for weight
to 0.47 for LDL-C whereas those for significant parent-offspring varied from 0.09 for
fasting plasma glucose to 0.43 for baPWV.
Conclusions: This study demonstrated significant heritability and familial
aggregation of cardiovascular risk factors in a random sample of ethnic Chinese
population.
Key words: heritability; cardiovascular risk factors; family study; highly sensitive
C-reactive protein; brachial-ankle pulse wave velocity
2
1. Introduction
Cerebrovascular disease (CVD), cardiovascular heart disease (CHD), and
diabetes are the most prominent causes of death worldwide. In Taiwan, CHD, CVD,
and diabetes were the second, third, and fourth leading causes of death in men and
women in 2011, accounting for more than 22.3% and 26.6% of total deaths,
respectively [1]. Cardiovascular risk factors associated with an increased risk of
developing CHD that tend to cluster in individuals include abdominal obesity, high
blood pressure, a low level of high-density lipoprotein cholesterol (HDL-C), a high
triglyceride (TG) level, and a high plasma glucose concentration [2, 3]. These
associated risk factors are referred to as the metabolic syndrome (MetS). Insulin
resistance due to obesity or an inherited genetic defect was hypothesized as the
mechanism underlying MetS.
MetS has been shown to increase the risk of CVD mortality and all-cause
mortality [3, 4]. The role of environmental factors in the aetiology of CVD is well
established. Inherited susceptibility to CVD is one of a number of possible risk factors
that could account for this finding. Previous studies investigating genetic factors of
cardiovascular risk factors include familial aggregation study [5-19], candidate gene
study [20], and GWA study [20-24]. Studies of familial aggregation of MetS and its
components focused on correlation of these phenotypes among family members
[9-12]. Most studies that reported heritability of cardiovascular risk factors based on
studies on twins either in white [15–17] or Asian populations [5–8 16]. Those studies
using twin study design in Asian populations either report cross-trait correlations or
diseases-discordant rates [5–8, 16]. Only one study reported heritability estimates of
C-reactive protein (CRP) and body fat composition markers in Chinese adolescent
twins [16]. Reported studies on estimates of heritability of cardiovascular risk factors
3
in Chinese using family study design are limited [12]. A previous study based on
Chinese adolescent probands and their relatives in a rural area reported that
heritability was highest in LDL (0.36), then total cholesterol (0.33), and systolic blood
pressure (0.32) [12]. Yet highly sensitive CRP (hs-CRP) and indicators for arterial
stiffness and peripheral vascular disease (PVD) in Chinese adults have not been
studied. Hs-CRP is a marker of systemic inflammation. Mild chronic elevations of
hs-CRP concentrations are independently predictive of future cardiovascular events
[17]. Arterial stiffness and PVD are two independent predictors of cardiovascular
mortality and morbidity [18].
One problem encountered when attempting to assess the importance of genetic
factors is the difficulty in determining how much of observed familial aggregation of
CVD cases can be explained by shared environmental factors and how much is due to
genetic predisposition. Failure to control several environmental risk factors while
examining familial clustering was a major weakness of many previous studies. These
factors include exposure to environmental tobacco smoke and physical activity, both
of which have been confirmed to be associated with cardiovascular risk factors. The
purpose of this study was to estimate heritability of cardiovascular risk factors in
relatives of residents who participated in the Taichung Community Health Study and
Family Cohort (TCHS-FC) while controlling as many potential confounders as
possible. If familial aggregation exists, this study would provide the rationale to
investigate both genetic and intra-familial environmental causes.
2. Methods
2.1 Participants
This community-based cross-sectional study was based on data from
participants of TCHS and their family members (TCHS family cohort, TCHS-FC).
4
Detailed TCHS methodology is described elsewhere [25]. Briefly, target population
consisted of residents aged 40 and over in Taichung City, Taiwan in October 2004. A
two-stage sampling design was used to draw residents, with a sampling rate
proportional to sample size within sampling units of each stage. At each stage, simple
random sampling was performed. A total of 3,530 subjects were eligible, and 2,359
agreed to participate with an overall response rate of 66.83%. The current study had a
total of 1564 study subjects. Among them, 484 were TCHS participants and 1080
were their family members. Considering that several measurements, such as hs-CRP
and insulin, were only determined in the first 1305 subjects of the TCHS participants,
these measurements were only available in 607 to 978 participants. Their ages were
between 12 and 91 years and 707 were males, which corresponds to a proportion of
45.2%. Mean age of men was 49.84 years and that of women was 48.88 years. The
1,564 subjects belonged to 494 families: 5 families of size 1, 181 families of size 2,
163 of size 3, 83 of size 4, 31 of size 5, and 31 more than size 6. Relationships
consisted of sibships (178 sib pairs), 641 parent-offspring pairs, and 9 avuncular pairs.
2.2 Measures
Laboratory examination
Blood was drawn from an antecubital vein in the morning after a 12-h overnight
fasting and was sent for analysis within 4 h of blood collection. Biochemical markers,
such as fasting plasma glucose, HDL-C, and triglyceride, were analyzed by a
biochemical autoanalyzer (Beckman Coulter Synchron system, Lx-20, Fullerton, CA,
USA) at the Clinical Laboratory Department of China Medical University Hospital.
Fasting plasma glucose was measured in blood obtained using NAF TUBE. NAF
5
TUBE contains 5mg sodium fluoride to inhibit glucose metabolism and 4mg
potassium oxalate to chelate calcium and to prevent coagulation. Inter- and intra-assay
coefficient of variations (CVs) for fasting plasma glucose were both 4%. We
measured cholesterol and triglyceride in serum mode. Triglyceride levels were
determined by an enzymatic colorimetric method. Inter- and intra-assay CVs for
triglyceride were 6.8% and 5%, respectively. HDL-C level was measured by a direct
HDL-C method and the interassay and intraassay CVs were both 4.5%. Low-density
lipoprotein cholesterol (LDL-C) level was also measured by a direct LDL-C method
and inter- and intra-assay CVs were 4.5% and 3%, respectively. Hs-CRP levels were
measured by nephelometry, a latex particle-enhanced immunoassay (TBA-200FR,
Tokyo, Japan). Inter- and intra-assay CVs were <2.0% and <1.9%, respectively.
Lower detection limit of the assay was 0.1 mg/L. Serum insulin level was measured
by a commercial enzyme-linked immunosorbent assay kit (Diagnostic Products, Los
Angeles, CA). Inter-assay CV for insulin was 8.7% and intra-assay CV was 3.4%.
Brachial-ankle pulse wave velocity (baPWV) and ankle-brachial index (ABI) were
determined non-invasively with subjects in supine position and cuffs wrapped on both
brachia and ankles with a VP-1000 automated PWV/ABI analyzer (PWV/ABI; Colin
Co. Ltd., Komaki, Japan). BaPWV is calculated as the distance traveled by pulse
wave from brachium to ankle divided by time taken to travel the distance. It has been
6
well-documented validity and reproducibility (coefficient of variation [CV]=3.31%
and reproducibility coefficient=0.947). For every subject, maximum of left and right
baPWV was chosen. Higher baPWV values indicated more severe arterial stiffness.
Lower ABI values indicated more severe PVD. Trained staff measured blood pressure
(BP) in the right arm using an appropriately sized cuff and a standard mercury
sphygmomanometer in a seated position. Physicians measured BP using the same
method in the same arm while they did the physical examination. If differences of BP
measured between trained staff and physicians exceed 5 mmHg (either systolic BP or
diastolic BP), then BP measurement was taken for a third time by the same physician.
Anthropometric measurement, demographic factors, and lifestyle behaviors
Anthropometric measurements were obtained from complete physical
examination. Weight and height were obtained on an autoanthropometer (super-view,
HW-666), with subjects shoeless and wearing light clothing. With participant standing,
waist circumference was measured midway between superior iliac crest and costal
margin by fitting the tape snugly but not compressing soft tissues. Hip circumference
was measured at the maximum protrusion point of buttocks around pelvis.
Circumference was measured to the nearest 0.1cm. Then, waist-to-hip ratio (WHR)
was calculated as a measure of regional fat distribution. WHR and waist-to height
ratio (WhtR) were calculated by dividing WC by hip circumference and body height,
respectively. WHtR corrects WC for an individual’s height and is another measure of
central obesity. Body mass index (BMI) was derived from the formula of weight (kg)
 (height)2 (m2). Percent body fat mass (%FM) was determined by BIA for total body,
7
which was assessed by a body composition analyzer (Tanita BC-418, Arlington
Heights, Illinois, USA). Amount of body fat was expressed as a percentage of total
weight. Previous studies confirmed validity of BIA in estimating body composition
compared with dual energy X-ray absorptiometry (DEXA) [26]. BIA was a good
predictor of DEXA-derived fat-free mass (r=0.85 to 0.88) and was superior to BMI in
measuring body fat [27]. It has been shown that BIA, as a simple method, has practice
value in measuring anthropometric attributes for estimating their heritability [27].
Information on smoking, alcohol drinking, physical activity, and betel nut
chewing were obtained by questionnaire when participants underwent a complete
physical examination. Smoking, alcohol drinking, physical activity, and betel nut
chewing were dichotomized into two groups. Those in non-smoking group did not
ever smoke or ever smoked less than 100 cigarettes during their lifetime, whereas
those in smoking group smoked currently or ever smoked greater than or equal to 100
cigarettes during their lifetime. Individuals who self-reported having alcohol drinking,
betel nut chewing, or physical activity were classified into the group with this specific
characteristic. Participants who were defined as having physical activity were those
who had self-reported that they had engaged in physical activities regularly at least
twice a week and had spent at least 20 minutes each time for 6 months.
2.3 Statistical analysis
Simple descriptive analyses, such as mean, standard deviation, and proportion, were
employed to analyze data when appropriate. Intraclass correlations for sibling pairs,
parent-offspring and spouse pairs were estimated for all cardiovascular risk factors. To
determine contribution of genetic factors to cardiovascular risk factors, heritability
estimates (h2) were calculated using variance components approach, which is
implemented in SOLAR [28]. To determine contribution of genetic factors to
8
cardiovascular risk factors, a quantitative trait, y, for individual i was modeled as
yi      j X ij  g i  ei , where  is trait mean, X ij is the j-th covariate with
regression coefficient  j , gi is an additive genetic effect normally distributed with
mean of 0 and variance of  2g , and ei is a random residual effect which is normally
distributed with a mean of 0 and variance of  e2 . Any non-additive genetic,
unmeasured measurement, and random error are incorporated into ei . Sum of  2g
and  e2 is equal to 1. Heritability is estimated by  2g , which is defined as relative
proportion of residual variance in the quantitative trait explained by additive genetic
factors divided by residual variance after adjustment for covariates. Maximum
likelihood methods were used to simultaneously estimate mean and variances as well
as the covariate and genetic effects. Significance of covariate effects was assessed by
likelihood ratio tests, which are used to assess significance of a parameter of interest
by comparing log-likelihood of the model in which parameter is estimated with that of
the model in which parameter is constrained to zero. Heritability estimates for
cardiovascular risk factors were calculated by adjusting for age, sex, smoking, alcohol
drinking, physical activity, and betel nut chewing. Estimates of covariate variance
obtained were used to estimate percentage of total variation explained by genetic
factors: [(1−proportion of variance explained by covariates) × h2] × 100.
3. Results
Demographic and laboratory characteristics are summarized in Table 1.
Estimates of heritability, proportion of variance attributed to covariates, and
proportion of variance attributed to genetic factors are shown in Table 2.
Cardiovascular risk factors with estimates of heritability greater than 40% after
multivariate adjustment were TG (h2=0.41; p<0.001), HDL-C (h2=0.49; p<0.001),
9
LDL-C (h2=0.47; p<0.001), total cholesterol (h2=0.46; p<0.001), hip circumference
(h2=0.44; p<0.001), and weight (h2=0.42; p<0.001). Cardiovascular risk factors with
estimates of heritability between 20% and 40% included insulin (h2=0.39; p<0.001),
SBP (h2=0.24; p<0.001), ABI (h2=0.25; p=0.0011), hs-CRP (h2=0.38; p=0.0011),
waist circumference (h2=0.26; p<0.001), BMI (h2=0.38; p<0.001), %FM (h2=0.35;
p<0.001), and WHtR (h2=0.24; p<0.001). Cardiovascular risk factors with more than
20% of variance attributed to all covariates included SBP (36.26%), baPWV (48.24%),
waist circumference (33.09%), %FM (39.81%), weight (33.46%), WHR (45.15%),
and WHtR (25.05%). Cardiovascular risk factors with more than 20% of variance
attributed to genetic factors included insulin (38.57%), TG (34.77%), HDL-C
(40.41%), LDL-C (46.63%), total cholesterol (44.58%), ABI (25.06%), hs-CRP
(35.88%), BMI (34.98%), %FM (21.20%), hip circumference (41.99%), and weight
(28.23%).
Table 3 shows results of significant covariates including age, gender, smoking,
alcohol drinking, physical activity, and betel nut chewing, which were associated with
cardiovascular risk factors. Age and gender were significant factors explaining
variation of all cardiovascular risk factors, except for age in insulin and gender in total
cholesterol. Smoking was associated with SBP, DBP, TG, HDL-C, and WHR; alcohol
drinking with FPG, insulin, HDL-C, and baPWV; physical activity with insulin, TG,
ABI, waist circumference, WHR, and WhtR; betel nut chewing with insulin, SBP,
DBP, TG, HDL-C, waist, BMI, %FM, WHR, and WHtR. These findings indicate that
environmental factors were also significant predictors of variation in cardiovascular
risk factors.
Table 4 shows intra-class correlation for spouse, sibling, and parent-offspring
relationship. Significant weak or moderate familial correlation coefficients for spouse
10
were HDL-C, baPWV, %FM, hip circumference, and weight. Cardiovascular risk
factors with significant and moderate sibling correlation coefficients (>0.2) were
HDL-C, LDL-C, total cholesterol, baPWV, %FM, hip, and WHtR. All
parent-offspring correlation coefficients were significant except for ABI. Those with
moderate correlation coefficients were insulin, TG, HDL-C, LDL-C, total cholesterol,
baPWV, and hip circumference.
4. Discussion
In our community-based sample of middle-aged adults and their family
members, we found that most cardiovascular risk factors are heritable. Results reveal
a significant and moderate magnitude of heritability for insulin, TG, HDL-C, LDL-C,
total cholesterol, hs-CRP, BMI, and hip circumference and a weak magnitude of
heritability for FPG, SBP, DBP, baPWV, and WHR. Although these cardiovascular
risk factors have a complex pathogenesis that is likely influenced by the interaction of
numerous environmental and genetic factors, evidence for genetic effects suggests a
substantial genetic component for variation in cardiovascular risk factors, even after
accounting for effects of lifestyle factors. Prior studies have demonstrated that genetic
factors contribute to variation in levels of cardiovascular disease-related variables,
including hypertension [21], diabetes [22], cholesterol levels [23], and measures of
obesity [24]. To our knowledge, our study using a random sampling method for
participant selection represent the first large population-based estimate of heritability
for a large number of cardiovascular risk factors in Chinese adults who resided in
community with integrated on-the-site screenings for hypertension, diabetes, and
hyperlipidemia for adults aged 40 years and over. Our results suggest that these
cardiovascular risk factors may be useful measures for further exploring genetics of
atherosclerosis and clinical cardiovascular disease.
11
Although several studies have reported familial correlation and heritability of
cardiovascular risk factors among different populations [9–15, 27, 29, 30], little has
been performed on Chinese population [9, 12, 30]. Chien et al. [12] estimated
heritability and familial correlation of parent-offspring and sibling on MetS-related
quantitative components on ethnic Chinese community based on adolescent probands
[12]. Feng et al. [9] reported sibling intra-class correlation in a Chinese population.
Wu et al. [30] estimated heritability of metabolic factors and insulin variables based
on siblings with Japanese or Chinese families with hypertension. Consistent with
findings of previous studies [9, 10, 12, 13, 15], we demonstrated the strongest familial
correlation and heritability in lipid-related quantitative components. On the contrary,
estimates of heritability and familiar correlation in our study were either higher or
lower than those estimated in other populations [9-13, 15, 27, 31, 32]. Compared with
studies conducted among Chinese, estimates of heritability and intra-class correlation
for parental offspring and sibling relationships in our study were close to those
estimated by Chien et al. [12] and Feng et al. [9]. Our study further provides estimates
of familial correlation and heritability of insulin, hs-CRP, baPWV, ABI, WHR, and
WHtR. We demonstrated moderate heritability (>30%) in insulin, ABI, hs-CRP, and
hip circumference; strong sibling correlation (>0.2) in baPWV, %FM, hip
circumference, and WHtR; and strong parent-offspring correlation in insulin, baPWV,
and hip circumference. Heritability of ABI, baPWV, and hs-CRP has not been
reported in Chinese adult population. With exception of sibling correlation of insulin
and %FM in a Chinese population reported by Feng et al. [9], these significant
familiar correlations are reported for the first time in the current paper. Our study
results strongly suggest genetic susceptibility in cardiovascular risk factors.
Our findings regarding heritability of lipid profiles are consistent with those
12
reported by Chien et al based on Chinese adolescent probands and their relatives in a
rural area [12], that estimated heritability was highest in lipid profiles. Results about
percentage of variance attributable to genetic factors is so much higher for lipids than
blood pressure are consistent with findings of previous studies which demonstrated so
many loci have been found for lipids [33] versus blood pressure phenotypes [34]. On
the contrary, our finding about high percentage of variance attributable to genetic
factors for lipids was inconsistent with those reported by Jermendy et al [35]. In
Jermendy’s study, 63 monozygotic and 38 dizygotic adult twin pairs were investigated
the effect of genetic and environmental influences on cardio-metabolic risk factors
and they found lipid profiles were low heritable, but shared and unique environmental
influences had the highest proportion of total phenotypic variance in lipid profiles.
The possible explanation for the differences may be due to our study did not consider
shared environmental influences.
This study had three advantages. First, study subjects were ascertained from a
representative sample of well-defined geographic population and distributions of
demographic factors are similar to our target population. Our estimates of heritability
and familiar correlation are thus more reflective of conditions in our population.
Heritability studies using twin design are based upon assumption that twins are
representative of general population for outcomes being studies. Although twin design
has advantage of allowing greater control over observable environmental risk factors,
it has several limitations. A major limitation is that twin study may restrict inferences
to general population because clusters themselves may not be representative. In
addition, twins reared together have a unique social and behavioral development
experience because they grow up with a co-twin, while twins reared apart are likely to
have been involved in unusual situations leading to their separation. Furthermore,
13
recruitment of twin is difficult and there is substantial self-selection. This may result
in differential participation rates that may lead to selection bias. Second, our study
provides intra-class correlation for spouses, siblings, and parent-offspring whereas
previous twin studies can only report intra-class correlation for siblings. Third, we
adjusted for many lifestyle behaviors, such as physical activity, smoking, alcohol
drinking, and betel nut chewing, some of which were neglected by many previous
studies. However, we did not consider dietary habits, which were important
determinants for cardiovascular risk factors.
Our study had several important limitations that must be stated. First, given
limited funding, some measurements were restricted to the first 1305 TCHS subjects
entering the current study, indicating that potential selection bias might exist. To
assess this possibility, we examined demographic characteristics of individuals with
and without these measurements by comparing age, sex, and administrative unit, and
similar distributions were found. Non-differential distributions in age, sex, and
administrative unit indicate that this type of selection error might be random. Thus,
biased results in the effect may be toward the null, a lesser threat to validity. Second,
the models we used assumed that there is no gene-covariate interaction, i.e.
heritability for a marker is constant across covariate spectrum such as age, and gender,
which is commonly adopted by most studies. However, previous studies indicate that
correlation among family members for lipids and blood pressure depended on age, i.e.,
gene-age interactions [36, 37].
Third, we did not consider passive smoking status, dietary factors, and status of
hypertension, diabetes, and hyperlipidemia or their medication in our analysis.
Passive smoking status was an important determinant for cardiovascular risk factors,
14
but we did collect such information. Data from Adult smoking behavior Surveillance
system in Taiwan reported the prevalence of passive smoking in Taiwan was 24.9% in
family and 15.7% in occupation setting in 2010 [38]. In Taiwan, a policy regarding
prohibiting smoking in public settings was implemented, thus we believed prevalence
of passive smoking would become lower. Due to low prevalence of passive smoking,
its confounding effect on our heritability estimates would be small. For dietary factors,
a twin study indicated that genetic variants in key leukotriene enzymes are associated
with atherosclerosis, and dietary intake of polyunsaturated fatty acids modifies the
effect of leukotriene variants on atherosclerosis [39, 40]. Furthermore, dietary factor
is also associated with a favorable change in distribution of HDL subspecies. Thus, no
considering dietary factors may have impact on estimates of heritability of baPWV,
ABI, and HDL-C in our study. Regarding that we also did not consider status of
hypertension, diabetes, and hyperlipidemia and their medication in our analysis, it is
because prevalence of these diseases’ medication is low (<20%) and their
confounding effects would be small. In addition, Wu et al. reported that similar results
of heritability of metabolic variables were found no matter antihypertensive
medication was considered or not [30]. Fourth, our sample was primarily Chinese,
limiting generalizability of our findings to other racial and ethnic groups. Studies of
racially diverse samples observed a high variation in prevalence of cardiovascular risk
15
factors among Blacks, Whites, Chinese, and South Asia [41]. Thus, heritability
estimates from our results may not be extended to other racial/ethnic groups or Asian
populations. Fifth, a limitation of measuring the brachial–ankle PWV is this
parameter reflects not only elastic arterial stiffness, but also muscular arterial stiffness
[42]. Even so, previous study demonstrated brachial–ankle PWV shows a close
correlation with aortic PWV [42] and carotid–femoral PWV [43], the gold standard
for assessing central arterial stiffness [44]. Due to a high level of skill and exposure of
inguinal region being required for carotid–femoral PWV measurement, so its
applicability is mostly limited to research institutes [45], not in clinical practice. On
the contrary, baPWV has been shown to be a promising technique to assess arterial
stiffness conveniently in clinical practice. As for ABI, it has been reported its
reliability was 0.61, lower than minimal requirement of 0.7. Thus, it was suggested
there was a need for repeated measures of ABI in clinical practice [46]. Sixth, we did
not consider shared environment and epigenetic factors. For shared environment,
heritability estimates in the current study may overestimate genetic contribution [47]
because we did not measure shared environments. However, 87.62% of siblings
indicated that they lived in separate households either from one another or from their
parents at the time of the study. Shared environments early in life may possibly
contribute to correlations for these cardiovascular risk factors among adult relatives.
16
For spouses, we cannot separate how much of their correlation is attributable to
shared environments and how much is attributable to assortative mating for factors
related to these cardiovascular risk factors. Spouse correlation coefficients for all
cardiovascular risk factors were much weaker than those of sibling and
parent-offspring, except for HDL-C and baPWV. We could assume that spouse
correlation is fully attributable to shared environments, which is likely, because of
their weak association. According to our findings, genetic factors would be stronger
factors that explain variation of these cardiovascular risk factors because of high
estimate of heritability, especially for lipid profiles and obesity markers. Recent
studies explored epigenetic effect on phenotypes using dizygotic and monozygotic
twin studies [48-50]. It has been shown that a difference in environment can produce
different developmental monozygotic twin subtypes with respective to their
susceptibility, resulting in discordant in autoimmune disease status [48]. In a
high-throughput study of epigenetic differences between monozygotic twins, they
found there is an age-dependent accumulation of epigenetic differences between two
siblings of twin pairs [49]. On the contrary, a epigenetic study of DNA methylation
status of about 6,000 unique genomic regions provide evidence that epigenetic
similarity contribute to phenotypic similarities in monozygotic co-twins [50]. Due to
no study examining the effect of epigenetic factors on heritability of cardiovascular
17
risk factors, it is hard to know the impact of epigenetic factors on our findings. In the
future, studies estimating heritability of cardiovascular risk factors may take into
account of epigenetic factors.
6. Conclusion
In conclusion, a low or modest to high proportion of variability in
cardiovascular risk factors is attributed to genetic factors. Further studies of genetic
linkage, candidate gene association, or GWAS association considering shared
environment factors are necessary to identify specific genetic variants associated with
this important predictor of cardiovascular disease events and mortality. These
cardiovascular risk factors are also associated with environment factors including
smoking, alcohol drinking, physical activity, or betel nut chewing.
18
Author contributions
CCL, PAP and TCL developed the hypotheses tested in the study. SLRK CCL,
CSL and WYL contributed to study design, planned and conducted the study,
performed the vascular measurements, developed diet materials, performed the
statistical analyses, interpreted the results and wrote the manuscript. TCL, PAP, CIL
and JSC contributed to statistical analyses, interpretation of the data and critically
reviewed the manuscript.
Disclosures
None.
Conflict of interest
The authors report no relationships that could be construed as a conflict of
interest.
Acknowledgements
This study was supported primarily by the Ministry of Science and Technology
of Taiwan (National Science Council) (NSC94-2314-B039-019,
NSC95-2314-B-039-009, NSC97-2314-B-039-019 & NSC99-2628-B039-007-MY3),
the China Medical University Hospital (DMR101-110), and Taiwan Ministry of
Health and Welfare Clinical Trial and Research Center of Excellence
(DOH102-TD-B-111-004).
19
Reference
1. Ten leading causes of death from Ministry of Health and Welfare, R.O.C:
http://www.mohw.gov.tw/EN/Ministry/Statistic.aspx?f_list_no=474.
2. Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol
in Adults. Executive summary of the third report of the National Cholesterol
Education Program (NCEP) expert panel on detection, evaluation, and treatment
of high blood cholesterol in adults (Adult Treatment Panel III). JAMA
2001;285:2486-97.
3. Isomaa B, Almgren P, Tuomi T, Forsén B, Lahti K, Nissén M, et al.
Cardiovascular morbidity and mortality associated with the metabolic syndrome.
Diabetes Care 2001;24:683-9.
4.
Klein BE, Klein R, Lee KE. Components of the metabolic syndrome and risk of
cardiovascular disease and diabetes in beaver dam. Diabetes Care
2002;25:1790-4.
5.
Yang H, Li X, Cao W, Lu J, Wang T, Zhan S, et al. Chinese National Twin
Registry as a resource for genetic epidemiologic studies of common and complex
diseases in China. Twin Res. 2002;5(5):347-51.
Lee K, Song YM, Sung J. Genetic and environmental associations between
C-reactive protein and components of the metabolic syndrome. Metab Syndr
Relat Disord. 2013;11(2):136-42.
Lee K, Sung J, Lee SC, Park SW, Kim YS, Lee JY, et al. Segment-specific
6.
7.
carotid intima-media thickness and cardiovascular risk factors in Koreans: the
Healthy Twin Study. Eur J Prev Cardiol. 2012;19(5):1161-72.
8. Li L, Gao W, Yu C, Lv J, Cao W, Zhan S, et al. The Chinese National Twin
Registry: an update. Twin Res Hum Genet. 2013;16(1):86-90.
9. Feng Y, Zang T, Xu X, Xu X. Familial aggregation of metabolic syndrome and
its components in a large Chinese population. Obesity (Silver Spring).
2008;16(1):125-9.
10. Lee KE, Klein BE, Klein R. Familial aggregation of components of the multiple
metabolic syndrome in the Framingham Heart and Offspring Cohorts: Genetic
Analysis Workshop Problem 1. BMC Genet. 2003;4 Suppl 1:S94.
11. Park HS, Park JY, Cho SI. Familial aggregation of the metabolic syndrome in
Korean families with adolescents. Atherosclerosis. 2006;186(1):215-21.
12. Chien KL, Hsu HC, Chen WJ, Chen MF, Su TC, Lee YT. Familial aggregation
of metabolic syndrome among the Chinese: report from the Chin-Shan
community family study. Diabetes Res Clin Pract. 2007;76(3):418-24.
20
13. Austin MA, King MC, Bawol RD, Hulley SB, Friedman GD. Risk factors for
coronary heart disease in adult female twins. Genetic heritability and shared
environmental influences. Am J Epidemiol. 1987;125(2):308-18.
14. Edwards KL, Newman B, Mayer E, Selby JV, Krauss RM, Austin MA.
Heritability of factors of the insulin resistance syndrome in women twins. Genet
Epidemiol. 1997;14(3):241-53.
15. Poulsen P, Vaag A, Kyvik K, Beck-Nielsen H. Genetic versus environmental
aetiology of the metabolic syndrome among male and female twins.
Diabetologia. 2001;44(5):537-43.
16. Wang G, Christoffel KK, Brickman WJ, Hong X, Arguelles L, Zhang S, et al.
C-reactive protein in adolescent twins: patterns and relationship to adiposity. J
Clin Endocrinol Metab. 2011;96(10):3226-33.
17. Ridker PM, Rifai N, Rose L, Buring JE, Cook NR. Comparison of C-reactive
protein and low-density lipoprotein cholesterol levels in the prediction of first
cardiovascular events. N Engl J Med 2002;347:1557-65.
18. Blacher J, Asmar R, Djane S, London GM, Safar ME. Aortic pulse wave velocity
as a marker of cardiovascular risk in hypertensive patients. Hypertension
1999;33:1111-7.
19. Edwards KL, Newman B, Mayer E, Selby JV, Krauss RM, Austin MA.
Heritability of factors of the insulin resistance syndrome in women twins. Genet
Epidemiol. 1997;14(3):241-53.
20. Gianfagna F, Cugino D, Santimone I, Iacoviello L. From candidate gene to
genome-wide association studies in cardiovascular disease. Thromb Res.
2012;129(3):320-4.
21. Levy D, DeStefano AL, Larson MG, O'Donnell CJ, Lifton RP, Gavras H, et al.
Evidence for a gene influencing blood pressure on chromosome 17. Genome
scan linkage results for longitudinal blood pressure phenotypes in subjects from
the Framingham Heart Study. Hypertension 2000;36:477-83.
22. Meigs JB, Panhuysen CI, Myers RH, Wilson PW, Cupples LA. A genome-wide
scan for loci linked to plasma levels of glucose and HbA(1c) in a
community-based sample of Caucasian pedigrees: The Framingham Offspring
Study. Diabetes 2002;51:833-40.
23. Shearman AM, Ordovas JM, Cupples LA, Schaefer EJ, Harmon MD, Shao Y, et
al. Evidence for a gene influencing the TG/HDL-C ratio on chromosome
7q32.3-qter: a genome-wide scan in the Framingham Study. Hum Mol Genet
2000;9:1315-20.
24. Atwood LD, Heard-Costa NL, Cupples LA, Jaquish CE, Wilson PW, D'Agostino
RB. Genomewide linkage analysis of body mass index across 28 years of the
21
Framingham Heart Study. Am J Hum Genet 2002;71:1044-50.
25. Lin CC, Liu CS, Lai MM, Li CI, Chen CC, Chang PC, et al. Metabolic syndrome
and its associated risk factors in a Taiwanese metropolitan adult population.
BMC Public Health 2007;7(1):739-43.
26. Roubenoff R. Applications of bioelectrical impedance analysis for body
composition to epidemiologic studies. Am J Clin Nur 1996;64(suppl):459S-62S.
27. Tarnokia AD, Tarnokia DL, Meddab E, Cotichini R, Stazi MA, Fagnani C, et al.
Bioimpedance analysis of body composition in an international twin cohort.
Obes Res Clin Pract 2012;8(3):e291-7.
28. Almasy L, Blangero J. Multipoint quantitative-trait linkage analysis in general
pedigrees. Am J Hum Genet. 1998;62:1198-211.
29. Murabito JM, Guo CY, Fox CS, D'Agostino RB. Heritability of the
ankle-brachial index: the Framingham Offspring study. Am J Epidemiol.
2006;164(10):963-8.
30. Wu KD, Hsiao CF, Ho LT, Sheu WH, Pei D, Chuang LM, et al. Clustering and
heritability of insulin resistance in Chinese and Japanese hypertensive families: a
Stanford-Asian Pacific Program in Hypertension and Insulin Resistance sibling
study. Hypertens Res. 2002;25(4):529-36.
31. Hong Y, Pedersen NL, Brismar K, de Faire U. Genetic and environmental
architecture of the features of the insulin-resistance syndrome. Am J Hum Genet.
1997;60(1):143-52.
32. Carmelli D, Cardon LR, Fabsitz R. Clustering of hypertension, diabetes, and
obesity in adult male twins: same genes or same environments? Am J Hum Genet.
1994;55(3):566-73.
33. Global Lipids Genetics Consortium, Willer CJ, Schmidt EM, Sengupta S, Peloso
GM, Gustafsson S, et al. Discovery and refinement of loci associated with lipid
levels. Nat Genet. 2013;45(11):1274-83.
34. Tragante V, Barnes MR, Ganesh SK, Lanktree MB, Guo W, Franceschini N, et
al. Gene-centric meta-analysis in 87,736 individuals of European ancestry
identifies multiple blood-pressure-related loci. Am J Hum Genet.
2014;94(3):349-60.
35. Jermendy G, Horváth T, Littvay L, Steinbach R, Jermendy AL, Tárnoki AD, et
al. Effect of genetic and environmental influences on cardiometabolic risk
factors: a twin study. Cardiovasc Diabetol. 2011;10:96.
36. Shi G, Gu CC, Kraja AT, Arnett DK, Myers RH, Pankow JS, et al. Genetic effect
on blood pressure is modulated by age: the Hypertension Genetic Epidemiology
Network Study. Hypertension. 2009;53(1):35-41.
37. Snieder H, van Doornen LJ, Boomsma DI. The age dependency of gene
22
expression for plasma lipids, lipoproteins, and apolipoproteins. Am J Hum Genet.
1997;60(3):638-50.
38. Adult Smoking Behavior Surveillance System from Health Promotion
Administration, Ministry of Health and Welfare, R.O.C:
http://tobacco.hpa.gov.tw/Show.aspx?MenuId=581.
39. Zhao J, Goldberg J, Vaccarino V. Leukotriene A4 hydrolase haplotype, diet and
atherosclerosis: a twin study. Atherosclerosis. 2013;226(1):238-44.
40. Bogl LH, Maranghi M, Rissanen A, Kaprio J, Taskinen MR, Pietiläinen KH.
Dietary omega-3 polyunsaturated fatty acid intake is related to a protective
high-density lipoprotein subspecies profile independent of genetic effects: a
monozygotic twin pair study. Atherosclerosis. 2011;219(2):880-6.
41. Chiu M, Austin PC, Manuel DG, Tu JV. Comparison of cardiovascular risk
profiles among ethnic groups using population health surveys between 1996 and
2007. CMAJ. 2010;182(8):E301-10.
42. Yamashina A, Tomiyama H, Takeda K, Tsuda H, Arai T, Hirose K, et al.
Validity, reproducibility, and clinical significance of noninvasive brachial-ankle
pulse wave velocity measurement. Hypertens Res. 2002;25(3):359-64.
43. Sugawara J, Hayashi K, Yokoi T, Cortez-Cooper MY, DeVan AE, Anton MA, et
al. Brachial-ankle pulse wave velocity: an index of central arterial stiffness? J
Hum Hypertens. 2005;19(5):401-6.
44. Adams MR, Robinson J, McCredie R, Seale JP, Sorensen KE, Deanfield JE, et al.
Smooth muscle dysfunction occurs independently of impaired
endothelium-dependent dilation in adults at risk of atherosclerosis. J Am Coll
Cardiol 1998;32:123-7.
45. Bonny A, Lacombe F, Yitemben M, Discazeaux B, Donetti J, Fahri P, et al. The
2007 ESH/ESC guidelines for the management of arterial hypertension. J
Hypertens. 2008;26(4):825.
46. Weatherley BD, Chambless LE, Heiss G, Catellier DJ, Ellison CR. The
reliability of the ankle-brachial index in the Atherosclerosis Risk in Communities
(ARIC) study and the NHLBI Family Heart Study (FHS). BMC Cardiovasc
Disord. 2006;6:7.
47. Tarnoki AD, Tarnoki DL, Bogl LH, Medda E, Fagnani C, Nisticò L, et al.
Association of body mass index with arterial stiffness and blood pressure
components: a twin study. Atherosclerosis. 2013;229(2):388-95.
48. Ballestar, E. Epigenetics lessons from twins: prospects for autoimmune disease.
Clinical Reviews in allergy & immunology 2009;39(1):30-41.
23
49. Fraga MF, Ballestar E, Paz MF, Ropero S, Setien F, Ballestar ML, et al.
Epigenetic differences arise during the lifetime of monozygotic twins. Proc Natl
Acad Sci U S A 2005;102:10604-9.
50. Kaminsky ZA, Tang T, Wang SC, Ptak C, Oh GH, Wong AH, et al. DNA
methylation profiles in monozygotic and dizygotic twins. Nat Genet
2009;41:240-5.
24