Experimental Gerontology 60 (2014) 8386

Contents lists available at ScienceDirect

Experimental Gerontology
journal homepage: www.elsevier.com/locate/expgero

Effect of genetic and non-genetic factors, including aging, on waist

circumference and BMI, and inter-indicator differences in
risk assessment
Ge Xin, Liu Shong, Liu Hui
College of Medical Laboratory, Dalian Medical University, Dalian 116044, China

a r t i c l e i n f o a b s t r a c t

Article history: Objective: To assess the effect of genetic and non-genetic factors on indicators derived from waist circumference
Received 24 May 2014 (WC) and body mass index (BMI) as well as inter-indicator differences in risk assessment age-related diseases
Received in revised form 29 September 2014 including diabetes mellitus, coronary heart disease and liver cancer.
Accepted 7 October 2014 Methods: Height, weight and WC were measured in 100 families (students and their two parents), 41 subjects
Available online 8 October 2014
with regular physical exercise routines, and 170 patients with diabetes mellitus, coronary heart disease or liver
Section Editor: Holly M. Brown-Borg
cancer. The BMI, waistheight ratio (WHtR) and waist circumference density index (WCDI) were calculated
for each subject.
Keywords: Results: BMI was less affected by genetic factors, while WHtR and WCDI were greatly affected by genetic factors as
Abdominal obesity revealed using multiple regression analysis. BMI, WHtR and WCDI were all sensitive to physical exercise accord-
Genetic ing to ROC analysis; among these factors, the most sensitive indicator was WHtR. However, ROC analysis demon-
Geriatric disease strated that WCDI was more effective than BMI and WHtR for assessing the risk of three diseases.
Conclusions: WCDI more accurately reects the roles of both genetic and non-genetic factors, including aging,
which can better predict disease.
2014 Elsevier Inc. All rights reserved.

1. Introduction
Obesity is a medical condition in which excess body fat accumulates
to the extent that it may have an adverse effect on health. Obesity in-
creases the likelihood of various diseases, particularly senile diseases,
such as heart disease (Cronin et al., 2013; Chrysant and Chrysant,
2013; Oboh and Adedeji, 2011) and type 2 diabetes (Kalra, 2013; Ye,
2013; Radzevi and Ostrauskas, 2013). Consequently, obesity has been
found to reduce life expectancy (Preston and Stokes, 2011; Singh
et al., 2011; Finkelstein et al., 2010). Obesity is most commonly caused
by a combination of excessive food energy intake, lack of physical activ-
ity, and genetic susceptibility. Similar to many other medical conditions,
obesity results from the interplay between genetic and environmental
factors (Manco and Dallapiccola, 2012; Gonzalez-Bulnes and Ovilo,
2012; Choi and Yoo, 2013). The percentage of obesity attributed to ge-
netics varies and is dependent on the population examined, which
ranges from 40% to 70% (Phan-Hug et al., 2012).
People are considered obese when their body mass index (BMI)
exceeds a specic value. BMI is closely related to both percentages
of body fat and total body fat. In addition, obesity has been further
Corresponding author.
E-mail addresses: liuhui60@sina.com, liuhui_60@163.com (L. Hui).
evaluated in terms of fat distribution via the waisthip ratio (Radzevi
and Ostrauskas, 2013), waistheight ratio (WHtR) (Radzevi and
Ostrauskas, 2013) and waist circumference density index (WCDI) (Zhao
et al., 2013). Abdominal obesity, which is also known as belly fat or cen-
tral obesity, is excessive fat around the stomach and abdomen. Metabolic
disorders associated with many diseases are more closely related with
abdominal obesity than general obesity (Page et al., 2009; Siavash et al.,
2008; Mi et al., 2013). Metabolic disorders with aging are also closely
related with abdominal obesity compared to general obesity (Zhao
et al., 2013). In other words, excessive waist circumference (WC) appears
to be more of a risk factor for senile disease compared to BMI.
Because WC and BMI are affected by both genetic and environmental
factors and differ in their prediction of disease, it is necessary to under-
stand the differences and characteristics of how WC and BMI are affect-
ed by genetic and environmental factors. Combining the differences and
characteristics of WC and BMI in disease assessment will help to eluci-
date the functional characteristics of genetic and environmental factors
in the occurrence of specic diseases, with the ultimate objective of
targeted prevention of specic diseases by utilizing simple biological
markers, such as WC or BMI. In this article, we systematically compared
the effect of genetic factors and exercise on BMI, WHtR and WCDI, and
assessed the effect of BMI, WHtR and WCDI on diabetes, cardiovascular
disease and liver cancer. We found that some indicators sensitive to

http://dx.doi.org/10.1016/j.exger.2014.10.005
0531-5565/ 2014 Elsevier Inc. All rights reserved.
84 G. Xin et al. / Experimental Gerontology 60 (2014) 8386
external factors did not have a good assessment effect of disease risk,
and thus, choosing an appropriate indicator has a great signicance in
clinical application.
2. Subjects and methods
2.1. Subjects
We randomly selected 100 Chinese college students (independent
of height, weight or BMI) and their family members as genealogy re-
search subjects: of the students, 50 males had a median age of 20.6
1.56 and 50 females had a median age of 19.8 1.5. To obtain a repre-
sentation population, the study population was 50% males and 50%
females. The eligibility criteria included subjects whose parents were
both alive, students who did not report any specic exercise habits,
and students whose body shape had no signicant changes within
1 year.
As the observation group was affected by an extrinsic factor, we ran-
domly selected 41 morning exercisers (mainly for calisthenics, 21 males
and 20 females, median age of 61.1 2.7). The eligibility criteria includ-
ed self-reports of no diabetes, cardiovascular disease, hypertension and
other common chronic diseases; a yearly exercise time of more than
10 months; and a history of regular daily exercise for more than
1 year. These 41 subjects were recruited from the community and
were randomly selected as the control group; their age and gender
were matched with members of the observation group. The eligibility
criteria included a habit of not exercising; other characteristics were
the same for the observation group.
Patients enrolled in the Second Afliated Hospital of Dalian Medical
University, Dalian, Liaoning, People's Republic of China for diabetes,
coronary heart disease or liver cancer were selected for placement in
the disease observation group.
For type 2 diabetes, age stratication was adopted (4050 years,
5060 years, 6070 years; each stratum had 20 persons, 10 males
and 10 females). Thus, 60 patients with type 2 diabetes were
selected, according to these diagnostic criteria: 1) glycated hemoglo-
bin A1c 6.5%; 2) fasting plasma glucose 7.0 mmol/l; or 3) blood
glucose at 2 h in the oral glucose tolerance test 11.1 mmol/l. Exclu-
sion criteria included patients with complication of heart disease or
liver cancer. We randomly selected 60 persons who were randomly
selected from a health check-up center as the control group; their
age and gender corresponded one-to-one with subjects in the type
2 diabetes group.
For coronary heart disease, age stratication was adopted
(4050 years, 5060 years, 6070 years; each stratum had 20 persons,
10 males and 10 females). Thus, 60 patients with coronary heart disease
met the following diagnostic criteria: typical angina symptoms and aux-
iliary examination according to 1) normal electrocardiogram: a tempo-
rary T wave inversion or depression of the ST segment during an angina
attack; 2) treadmill exercise test; 3) myocardial perfusion imaging;
4) coronary artery computed tomography angiography; 5) dynamic
electrocardiogram; or 6) echocardiography. Exclusion criteria included
patients with complication of diabetes or liver cancer. We randomly
selected 60 persons from a health check-up center as the control
group, whose age and gender corresponded one-to-one with patients
in the coronary heart disease group.
The liver cancer group consisted of 50 subjects (56.8 7.5 years)
with liver cancer who had been diagnosed by surgical intervention.
Exclusion criteria included patients with complication of diabetes or
heart disease. The control group consisted of healthy gender- and age-
matched subjects compared to members of the cancer group.
The Institutional Ethics Committee of Dalian Medical University
approved the study and waived the need for a written informed con-
sent from the participants due to the observational nature of the
study.
2.2. Measurements
The height and weight of each subject were measured using scales
with precisions of 0.1 cm and 0.1 kg, respectively. The WC was mea-
sured using a measuring tape at the umbilical level. The height, weight
and WC of the subjects' parents in the genealogy research group were
measured under instructional guidance given over the telephone.
Patients with liver cancer were measured prior to the surgical
interventions.
2.3. Calculation of BMI, WHtR and WCDI
The BMI was obtained by dividing the person's weight in kilograms
by the square of his or her height in meters. The WHtR is the ratio of
the circumference of the waist to height.
The WCDI was calculated according to our previous published study
(Zhao et al., 2013). The WCDI is dened as the ratio of weight and body
volume based on WC; which was human density obtained from WC and
reected the WC in the absence of the effect of height and weight on
WC. The WCDI was obtained using the following equation (Zhao et al.,
2013):
 
2
WCDI W= P  H
0:5
;
where W represents weight (kg), H represents height (m), and P repre-
sents WC (m). A small WCDI were indicative of a high WC.
2.4. Statistical analysis
In the genealogy research group, the measurement indicators of
children were used as the dependent variable, while those of parents
were used as the independent variable. Multiple regression analyses
were performed. When the multiple regression equation was statistical-
ly signicant, the indicator was affected by genetic factors, in which the
degree was measured with R2, and was positively proportional to the
effect of genetic factors.
ROC analysis was performed in the exercise and control groups
to observe the effect of non-genetic factors on measurement
indicators, in which the degree was measured using area-under-the-
curve (AUC). If the AUC was 0.5, then there was no signicance. If the
AUC was b 0.5, then it was transformed using the following equation
to allow comparison:
AUCnew 1AUCoriginal :
ROC analysis was performed in each disease group to observe the
assessment effect of the measurement indicators on disease risk in
which effectiveness was measured using the AUC. The greater the
value, the better is the effect of assessment.
ROC analysis was performed between the disease group (in which
three diseases were combined into one group) and control group
(college students) to observe the optimal cut-off value.
Statistical analyses were performed using SPSS statistical analysis
software for Windows (SPSS, Chicago, IL, USA). A difference was consid-
ered to be statistically signicant when the p-value was less than 0.05
(two-tailed test).
Table 1
Raw anthropometric data of the genealogy of research subjects.
Indicators Children Father Mother
Men Women
Height (cm) 174.8 5.7 163.2 4.7 172.5 4.9 161.5 4.7
Weight (kg) 64.7 9.0 54.1 6.2 71.6 9.1 58.5 7.0
WC (cm) 79.4 7.2 71.6 5.0 86.5 7.4 75.8 6.6
 G. Xin et al. / Experimental Gerontology 60 (2014) 8386 85

3. Results Table 3
ROC analyses between the exercise and control groups.

Raw data of the genealogy research subjects are shown in Table 1. Indicators Group p-Value AUC
The regression equations of measurement indicators between children Physical activity Control
and parents are shown in Table 2. The multiple regression equation
BMI 21.38 1.93 25.91 3.32 b0.0001 0.879
was statistically signicant for WHtR and WCDI.
WHtR 0.442 0.024 0.519 0.040 b0.0001 0.949
The ROC analyses between the exercise group and control group are WCDI 84.38 7.05 74.73 6.02 b0.0001 0.864
shown in Table 3. The AUCs were greater than 0.85 for all three indica-
tors (BMI, WHtR and WCDI).
The ROC analyses between each disease group are shown in Table 4. adiposis hepatica, which can develop into liver cancer (Chen et al.,
For the three diseases, the AUC of BMI was 0.5590.655; the AUC of 2012; Tanaka et al., 2012; Shen et al., 2012); thus, we focused on
WHtR was 0.5460.868; and the AUC of WCDI was 0.7760.992. The these three diseases. Unexpectedly, WHtR did not perform the best in
WCDI was more effective than BMI and WHtR for assessing disease risk. the assessment of the above three diseases. WCDI was more effective
ROC analysis of WCDI between the disease group and control group compared to BMI and WHtR in assessing disease risk (Table 3). Impor-
(college students) is shown in Fig. 1. The optimal cut-off value of WCDI tantly, the liver cancer group had a BMI median that was smaller than
under the ROC curve (AUC = 0.860) was 76.0. the control group, whereas the WC was relatively large in the liver
cancer group as assessed using the WCDI, which further indicated that
4. Discussion the WCDI is a more effective indicator compared to BMI in the risk
assessment of diseases.
The height, weight and WC were simple to measure; various indica- These results suggested that genetic and non-genetic factors might
tors derived from height, weight and WC can reect the risk of many pose different disease risks, and an indicator that precisely reects
chronic diseases and thus be used as signs for health assessment. The these two types of factors might be most optimal in predicting disease
waisthip ratio is also an important indicator to evaluate abdominal risk (Prudente et al., 2012). WHtR was more sensitive to environmental
obesity (Carmienke et al., 2013; Qin et al., 2010); however, hip mea- factors, which might obscure the indication of genetic factors, while
surement is more prone to errors compared to the measurement of WCDI, which perfectly reects the role of genetic and non-genetic
height and weight. Accurate data on the hip can be obtained by measur- factors, demonstrated a better prediction of disease. Compared with
ing the bare buttocks, which the study subjects frequently do not accept. WHtR, WCDI also comprehensively assessed the role of height, weight
Thus, we focused on determining the BMI, WHtR and WCDI in this and aging factors (Zhao et al., 2013), which may explain why WCDI
study. These results revealed that BMI was less affected by genetic had a better assessment effect compared to WHtR.
factors in an assessment of a random population, while WHtR and BMI is still an important indicator for evaluating the risk of obesity
WCDI, which are indicators that both reect the WC, were greatly and disease occurrence; however, according to our results, BMI reects
affected by genetic factors. If genetic factors affecting WHtR and WCDI mainly environmental factors. A survey found that most patients had
are linked with specic diseases, then measurement of the WHtR or corrected their unhealthy behaviors to a specic degree after the onset
WCDI could reect the genetic risk of disease. of diseases, which is usually monitored with BMI. Thus, these ndings
Genealogy research is the comprehensive observation on a group of might explain the poor capacity to distinguish BMI in this study; howev-
genes on a polygenic phenomenon. Although a few genes have shown a er, our data showed that WHtR and WCDI, which reects genetic predis-
proved correlation with BMI (Yang et al., 2014; Qi et al., 2014; Gupta position, could appropriately evaluate the genetic risk of diabetes and
et al., 2012) or WC (Gupta et al., 2012; Surakka et al., 2011; Mutschler coronary heart disease.
et al., 2013; Velez Edwards et al., 2013; Xin et al., 2014), a negative nd- Taken together, the results of this study suggested that WCDI was
ing is still possible in a family study if the effect of the relevant gene was the best indicator for the health assessment of a random population. A
insufciently strong. Although the sample size in this study was rela- WCDI N76.0 indicates a lower occurrence risk of diseases, whereas a
tively small, it was acceptable for purposes of comparing the degree of WCDI b 76.0 indicates a higher occurrence risk of diseases, as assessed
inuence of genetic factors between BMI and WC. In this study, the using ROC analyses. Thus, WCDI has a practical signicance in clinical
BMI had not signicantly regressed among family members, which application within a reference range. On the basis of the results obtain-
suggested that the degree of inuence of genetic factors was relatively ed, the measurement of WCDI is recommended for the assessment of
smaller in BMI compared to WC. health status and disease risk in real time.
The occurrence of diseases is also associated with environmental fac- Although WCDI was closely associated with genetic relationships, it
tors. It is generally believed that the factor that affects BMI, WHtR and can be controlled by lifestyle changes. The WCDI averaged more than 84
WCDI in a pathological direction is also a risk factor for disease for subjects who often performed calisthenics; if the WCDI was
occurrence, which underlies the use of BMI, WHtR and WCDI in evaluat-
ing disease risk. Thus, a selected indicator sensitive to environmental fac-
tors may be a more accurate indicator for disease risk assessment. Our Table 4
ROC analyses between the disease and control groups.
results showed that BMI, WHtR and WCDI were all sensitive to physical
exercise, and of these factors, the most sensitive indicator was WHtR. Disease Group Indicators
Type 2 diabetes mellitus and coronary heart disease are diseases that BMI WHtR WCDI
are typically related to obesity. Obesity can also frequently result in
Diabetes mellitus Disease 26.72 3.07 0.58 0.04 62.56 6.95
Control 25.75 3.22 0.52 0.04 75.45 5.82
p-Value 0.096 b0.0001 b0.0001
Table 2 AUC 0.589 0.868 0.922
Regression equation of three indicators between children and parents. Coronary heart disease Disease 26.34 3.65 0.56 0.05 65.49 7.90
Control 25.75 3.22 0.52 0.04 75.45 5.82
Indicators Regression equation R2 F p-Value
p-Value 0.349 b0.0001 b0.0001
BMI Y = 21.425 + 0.108 father 0.146 0.028 1.387 0.255 AUC 0.559 0.733 0.844
mother Liver cancer Disease 24.40 3.20 0.53 0.04 67.33 7.65
WHtR Y = 0.278 + 0.462 father 0.134 mother 0.251 16.239 b0.0001 Control 26.21 3.23 0.52 0.04 74.60 5.34
WCDI Y = 34.149 + 0.457 father + 0.163 0.213 13.123 b0.0001 p-Value 0.006 0.439 b0.0001
mother AUC 0.655 0.546 0.776
86 G. Xin et al. / Experimental Gerontology 60 (2014) 8386
Fig. 1. WCDI in the disease and control groups. The optimal cutoff value of WCDI was 76.0.
controlled to above 76.0, then the disease risk was signicantly reduced.
Thus, future studies should assess and compare how much lifestyle
changes can reduce disease incidence or delay the disease process by
controlling the BMI, WHtR and WCDI.
Competing interests
None declared.
References
Carmienke, S., Freitag, M.H., Pischon, T., Schlattmann, P., Fankhaenel, T., Goebel, H.,
Gensichen, J., 2013. General and abdominal obesity parameters and their combina-
tion in relation to mortality: a systematic review and meta-regression analysis. Eur.
J. Clin. Nutr. 67 (6), 573585.
Chen, Y., Wang, X., Wang, J., Yan, Z., Luo, J., 2012. Excess body weight and the risk of
primary liver cancer: an updated meta-analysis of prospective studies. Eur. J. Cancer
48 (14), 21372145.
Choi, J.H., Yoo, H.W., 2013. Control of puberty: genetics, endocrinology, and environment.
Curr. Opin. Endocrinol. Diabetes Obes. 20 (1), 6268.
Chrysant, S.G., Chrysant, G.S., 2013. New insights into the true nature of the obesity
paradox and the lower cardiovascular risk. J. Am. Soc. Hypertens. 7 (1), 8594.
Cronin, O., Morris, D.R., Walker, P.J., Golledge, J., 2013. The association of obesity with
cardiovascular events in patients with peripheral artery disease. Atherosclerosis
228 (2), 316323.
Finkelstein, E.A., Brown, D.S., Wrage, L.A., Allaire, B.T., Hoerger, T.J., 2010. Individual and
aggregate years-of-life-lost associated with overweight and obesity. Obesity (Silver
Spring) 18 (2), 333339.
Gonzalez-Bulnes, A., Ovilo, C., 2012. Genetic basis, nutritional challenges and adaptive
responses in the prenatal origin of obesity and type-2 diabetes. Curr. Diabetes Rev.
8 (2), 144154.
Gupta, V., Khadgawat, R., Ng, H.K., Walia, G.K., Kalla, L., Rao, V.R., Sachdeva, M.P., 2012. As-
sociation of TCF7L2 and ADIPOQ with body mass index, waisthip ratio, and systolic
blood pressure in an endogamous ethnic group of India. Genet. Test. Mol. Biomarkers
16 (8), 948951.
Kalra, S., 2013. Diabesity. J. Pak. Med. Assoc. 63 (4), 532534.
Manco, M., Dallapiccola, B., 2012. Genetics of pediatric obesity. Pediatrics 130 (1),
123133.
Mi, S.Q., Yin, P., Hu, N., Li, J.H., Chen, X.R., Chen, B., Yan, L.X., Zhao, W.H., 2013. BMI, WC,
WHtR, VFI and BFI: which indictor is the most efcient screening index on type 2
diabetes in Chinese community population. Biomed. Environ. Sci. 26 (6), 485491.
Mutschler, J., Abbruzzese, E., Wiedemann, K., von der Goltz, C., Dinter, C., Mobascher, A.,
Thiele, H., Diaz-Lacava, A., Dahmen, N., Gallinat, J., Majic, T., Petrovsky, N., Thuerauf,
N., Kornhuber, J., Grnder, G., Rademacher, L., Brinkmeyer, J., Wienker, T., Wagner,
M., Winterer, G., Kiefer, F., 2013. Functional polymorphism in the neuropeptide Y
gene promoter (rs16147) is associated with serum leptin levels and waisthip ratio
in women. Ann. Nutr. Metab. 62 (4), 271276.
Oboh, H.A., Adedeji, A.A., 2011. Correlation of waisthip-ratio and waistheight-ratio to
cardiovascular risks factors in a Nigerian population. Niger. Q. J. Hosp. Med. 21 (1),
1624.
Page, J.H., Rexrode, K.M., Hu, F., Albert, C.M., Chae, C.U., Manson, J.E., 2009. Waistheight
ratio as a predictor of coronary heart disease among women. Epidemiology 20 (3),
361366.
Phan-Hug, F., Beckmann, J.S., Jacquemont, S., 2012. Genetic testing in patients with obesi-
ty. Best Pract. Res. Clin. Endocrinol. Metab. 26 (2), 133143.
Preston, S.H., Stokes, A., 2011. Contribution of obesity to international differences in life
expectancy. Am. J. Public Health 101 (11), 21372143.
Prudente, S., Dallapiccola, B., Pellegrini, F., Doria, A., Trischitta, V., 2012. Genetic prediction
of common diseases. Still no help for the clinical diabetologist! Nutr. Metab.
Cardiovasc. Dis. 22 (11), 929936.
Qi, Q., Chu, A.Y., Kang, J.H., Huang, J., Rose, L.M., Jensen, M.K., Liang, L., Curhan, G.C.,
Pasquale, L.R., Wiggs, J.L., De Vivo, I., Chan, A.T., Choi, H.K., Tamimi, R.M., Ridker,
P.M., Hunter, D.J., Willett, W.C., Rimm, E.B., Chasman, D.I., Hu, F.B., Qi, L., 2014. Fried
food consumption, genetic risk, and body mass index: gene-diet interaction analysis
in three US cohort studies. BMJ 348, g1610.
Qin, L., Corpeleijn, E., Jiang, C., Thomas, G.N., Schooling, C.M., Zhang, W., Cheng, K.K.,
Leung, G.M., Stolk, R.P., Lam, T.H., 2010. Physical activity, adiposity, and diabetes
risk in middle-aged and older Chinese population: the Guangzhou Biobank Cohort
Study. Diabetes Care 33 (11), 23422348.
Radzevi, L., Ostrauskas, R., 2013. Body mass index, waist circumference, waisthip ratio,
waistheight ratio and risk for type 2 diabetes in women: a casecontrol study. Public
Health 127 (3), 241246.
Shen, C., Zhao, C.Y., Zhang, R., Qiao, L., 2012. Obesity-related hepatocellular carcinoma:
roles of risk factors altered in obesity. Front. Biosci. (Landmark Ed.) 17, 23562370.
Siavash, M., Sadeghi, M., Salarifar, F., Amini, M., Shojaee-Moradie, F., 2008. Comparison of
body mass index and waist/height ratio in predicting denite coronary artery disease.
Ann. Nutr. Metab. 53 (34), 162166.
Singh, P.N., Haddad, E., Tonstad, S., Fraser, G.E., 2011. Does excess body fat maintained
after the seventh decade decrease life expectancy? J. Am. Geriatr. Soc. 59 (6),
10031011.
Surakka, I., Isaacs, A., Karssen, L.C., Laurila, P.P., Middelberg, R.P., Tikkanen, E., Ried, J.S.,
Lamina, C., Mangino, M., Igl, W., Hottenga, J.J., Lagou, V., van der Harst, P., Mateo
Leach, I., Esko, T., Kutalik, Z., Wainwright, N.W., Struchalin, M.V., Sarin, A.P., Kangas,
A.J., Viikari, J.S., Perola, M., Rantanen, T., Petersen, A.K., Soininen, P., Johansson, A.,
Soranzo, N., Heath, A.C., Papamarkou, T., Prokopenko, I., Tnjes, A., Kronenberg, F.,
Dring, A., Rivadeneira, F., Montgomery, G.W., Whiteld, J.B., Khnen, M.,
Lehtimki, T., Freimer, N.B., Willemsen, G., de Geus, E.J., Palotie, A., Sandhu, M.S.,
Waterworth, D.M., Metspalu, A., Stumvoll, M., Uitterlinden, A.G., Jula, A., Navis, G.,
Wijmenga, C., Wolffenbuttel, B.H., Taskinen, M.R., Ala-Korpela, M., Kaprio, J., Kyvik,
K.O., Boomsma, D.I., Pedersen, N.L., Gyllensten, U., Wilson, J.F., Rudan, I., Campbell,
H., Pramstaller, P.P., Spector, T.D., Witteman, J.C., Eriksson, J.G., Salomaa, V., Oostra,
B.A., Raitakari, O.T., Wichmann, H.E., Gieger, C., Jrvelin, M.R., Martin, N.G., Hofman,
A., McCarthy, M.I., Peltonen, L., van Duijn, C.M., Aulchenko, Y.S., Ripatti, S., 2011. EN-
GAGE consortium. A genome-wide screen for interactions reveals a new locus on
4p15 modifying the effect of waist-to-hip ratio on total cholesterol. PLoS Genet. 7
(10), e1002333.
Tanaka, K., Tsuji, I., Tamakoshi, A., Matsuo, K., Ito, H., Wakai, K., Nagata, C., Mizoue, T.,
Sasazuki, S., Inoue, M., Tsugane, S., Research Group for the Development and
Evaluation of Cancer Prevention Strategies in Japan, 2012. Obesity and liver cancer
risk: an evaluation based on a systematic review of epidemiologic evidence among
the Japanese population. Jpn. J. Clin. Oncol. 42 (3), 212221.
Velez Edwards, D.R., Naj, A.C., Monda, K., North, K.E., Neuhouser, M., Magvanjav, O.,
Kusimo, I., Vitolins, M.Z., Manson, J.E., O'Sullivan, M.J., Rampersaud, E., Edwards, T.L.,
2013. Gene-environment interactions and obesity traits among postmenopausal
AfricanAmerican and Hispanic women in the Women's Health Initiative SHARe
Study. Hum. Genet. 132 (3), 323336.
Xin, G., Yang, G., Hui, L., 2014. Study to assess whether waist circumference and changes
in serum glucose and lipid prole are independent variables for the CETP gene.
Diabetes Res. Clin. Pract. 106 (1), 95100.
Yang, F., Chen, X.D., Tan, L.J., Shen, J., Li, D.Y., Zhang, F., Sha, B.Y., Deng, H.W., 2014.
Genome wide association study: searching for genes underlying body mass index
in the Chinese. Biomed. Environ. Sci. 27 (5), 360370.
Ye, J., 2013. Mechanisms of insulin resistance in obesity. Front. Med. 7 (1), 1424.
Zhao, X., Zhu, S., Jia, X., Yu, L., Liu, H., 2013. Constructing a waist circumference density
index to predict biological age and evaluating the clinical signicance of waist
circumference density age. Exp. Gerontol. 48 (4), 422426.