1032

Serum uric acid is associated with metabolic risk

factors for cardiovascular disease in the Uygur
population
Nan F. Li, Hong M. Wang, Jin Yang, Ling Zhou, Xiao G. Yao, and Jing Hong
Appl. Physiol. Nutr. Metab. Downloaded from www.nrcresearchpress.com by TEXAS STATE UNIV on 08/09/13

Abstract: The prevalence of hyperuricemia is low in Uygurs, who have a high prevalence of cardiovascular risk factors
such as hypertension, overweight–obesity, dyslipidemia, hyperglycemia, and insulin resistance (IR). This study sought to
investigate the relationships between serum uric acid (UA) and these risk factors in this population. A cross-sectional study
was conducted in Uygurs (859 males, 1268 females) aged 20 to 70 years. Demographic data, systolic blood pressure
(SBP), diastolic blood pressure (DBP), body mass index (BMI), and fasting and postprandial blood were obtained, and bio-
logical measurements were determined. The mean of BMI, SBP, DBP, total cholesterol, high-density lipoprotein choles-
terol (HDL-c), low-density lipoprotein cholesterol (LDL-c), triglycerides, fasting blood glucose, fasting insulin, and
homeostasis model assessment insulin resistance index (HOMA-IR), and the prevalence of hypertension, IR, hyperglyce-
mia, overweight–obesity, hypercholesteremia, hyper-LDL-c, and hypertriglyceridemia increased with UA but the preva-
lence of hypo-HDL-c decreased (p < 0.05). Logistic regression analysis showed that the odds ratios for IR, overweight–
obesity, hypercholesteremia, hyper-LDL-c, and hypertriglyceridemia against the lowest UA group increased but decreased
for hypo-HDL-c (p < 0.05). The UA in the hypo-HDL-c group was lower than that of the controls; the prevalence of
hypo-HDL-c in hyperuricemia subjects was lower than in those with normal UA (p < 0.05). But the opposite results were
observed between overweight–obesity, hyperglycemia, IR, hypercholesteremia, hypertriglyceridemia, and hyper-LDL-c and
correspondence controls, respectively (p < 0.05). In Uygur, elevated UA is associated with overweight–obesity, hypercho-
For personal use only.

lesteremia, hyper-LDL-c, hypertriglyceridemia, hyperglycemia, and IR. The HDL-c level significantly increases with UA,
whereas the prevalence of hypo-HDL-c decreases. Further studies are needed to clarify why UA is positively correlated to
HDL-c.
Key words: serum uric acid, overweight–obesity, dyslipidemia, insulin resistance, hyperglycemia, hypertension.
Résumé : Chez les Uygurs, la prévalence d’hyperuricémie est faible tandis que la prévalence des facteurs de risque [hy-
pertension, surpoids/obésité, dyslipémie, hyperglycémie, insulinorésistance (IR)] est élevée. Cette étude se propose d’ana-
lyser chez les Uygurs la relation entre le taux sérique d’acide urique (UA) et les facteurs de risque. L’étude, de nature
transversale, porte sur 1268 femmes et 859 hommes âgés de 20 à 70 ans. On enregistre les données démographiques et on
évalue les pressions systolique (SBP) et diastolique (DBP), l’indice de masse corporelle (BMI); de plus, on prélève des
échantillons de sang en condition de jeûne et dans un état postprandial afin d’analyser la concentration des mesures biolo-
giques choisies. Les moyennes respectives de BMI, des SBP et DBP, du cholestérol total, des lipoprotéines à haute densité
(HDL-c), des lipoprotéines à faible densité (LDL-c), des triglycérides, du glucose sanguin à jeun, de l’insuline à jeun, du
HOMA-IR et la prévalence de l’hypertension, de l’IR, de l’hyperglycémie, du surpoids–obésité, de l’hypercholestérolémie,
de l’hyper-LDL-c et de l’hypertriglycéridémie augmentent avec la concentration d’UA, mais la prévalence de l’hypo-
HDL-c diminue (p < 0,05). D’après l’analyse de régression logistique, les rapports des cotes de l’IR, du surpoids–obésité,
de l’hypercholestérolémie, de l’hyper-LDL-c et de l’hypertriglycéridémie augmentent relativement au groupe présentant la
plus faible concentration d’UA, mais diminuent en présence d’hypo-HDL-c (p < 0,05). Le taux d’UA dans le groupe pré-
sentant une hypo-HDL-c est plus faible que dans le groupe de contrôle; la prévalence de l’hypo-HDL-c chez les sujets hy-
peruricémiques est plus faible que chez les individus présentant un taux normal d’UA (p < 0,05). Néanmoins, on observe
le contraire en ce qui concerne le surpoids–obésité, l’hyperglycémie, l’IR, l’hypercholestérolémie, l’hypertriglycéridémie
en présence d’hyper-LDL-c comparativement aux valeurs du groupe de contrôle (p < 0,05). Chez les Uygurs, un taux élevé
d’UA est associé au surpoids–obésité, à l’hypercholestérolémie, à l’hyper-LDL-c, à l’hypertriglycéridémie, à l’hypergly-
cémie et à l’IR. Le taux d’HDL-c augmente significativement avec le taux d’UA et la prévalence de l’hypo-HDL-c dimi-
nue. Il faut faire d’autres études pour établir pourquoi le taux d’UA est positivement associé au taux d’HDL-c.

Received 16 February 2009. Accepted 5 August 2009. Published on the NRC Research Press Web site at apnm.nrc.ca on 13 November
2009.
N.F. Li,1 H.M. Wang, J. Yang, L. Zhou, X.G. Yao, and J. Hong. The Center of Hypertension of the People’s Hospital of Xinjiang
Uygur Autonomous Region; The Center of Diagnosis, Treatment and Research of Hypertension in Xinjiang, Urumqi, Xinjiang 830001,
China.
1Corresponding author (e-mail: lnanfang@yahoo.com.cn).

Appl. Physiol. Nutr. Metab. 34: 1032–1039 (2009) doi:10.1139/H09-101 Published by NRC Research Press
 Li et al.
Mots-clés : acide urique sérique, surpoids–obésité, dyslipémie, insulinorésistance, hyperglycémie, hypertension.
[Traduit par la Rédaction]
______________________________________________________________________________________
Introduction
Effective prevention of cardiovascular disease (CVD) re-
quires early detection and correction of predisposing condi-
Appl. Physiol. Nutr. Metab. Downloaded from www.nrcresearchpress.com by TEXAS STATE UNIV on 08/09/13
tions and risk factors in susceptible patients. Hypertension,
overweight–obesity, dyslipidemia, insulin resistance (IR),
and diabetes mellitus (DM) are common risk factors for
CVD and play a causal role in its pathogenesis (Tanuseputro
et al. 2003; Cooper et al. 2006). Besides these, elevated se-
rum uric acid (UA) is associated with increased incidence of
CVD, as well as mortality from it (Strasak et al. 2008).
Although the exact mechanism has not been clarified, epide-
miological and clinical evidence has demonstrated a close
relationship between hyperuricemia and these cardiovascular
risk factors and they occur together more frequently than ex-
pected by chance (Shao et al. 2007; Conen et al. 2004).
Moreover, higher morbidity in victims and mortality from
CVD are observed when these disorders occur together in
the same individual, so there may be value in treating hyper-
uricemia. Therefore, a more in-depth understanding of the
relationships between UA and these disorders may assist
For personal use only.
clinicians in treating and preventing more efficiently the on-
set of these disorders and the possible subsequent risk of
cardiovascular complications.
Previous study showed that the Uygur population, who
dwell in the Hetian area in the south of the Xinjiang Uygur
Autonomous Region of China, had a high prevalence of hy-
pertension (57.7%); obesity (68.2%); DM, impaired fasting
glucose (IFG), and impaired glucose tolerance (IGT)
(37.8%); hypo-high-density lipoprotein cholesterol (HDL-c)
(8.9%); and hypertriglyceridemia (25.3%) (Guo et al. 2006).
But the prevalence of hyperuricemia in Uygurs living in the
Hetian area was only 3.1%, which is far lower than that of
the Uygur population living in Urumqi (17.6%) (Sun et al.
2007) and the Han population in Beijing (15.4%) (Li et al.
1997), Qingdao (25.3%) (Nan et al. 2006), Chengdu
(13.2%) (Zhu et al. 2002), and Nanjing (13.3%) (Shao et al.
2003). The low prevalence of hyperuricemia and the high
prevalence of hypertension, overweight–obesity, dyslipide-
mia, hyperglycemia (DM, IFG, and IGT included), and IR
have displayed a so-called ‘‘separated phenomenon’’. This
study, then, focuses on whether the relationship between hy-
peruricemia and these cardiovascular risk factors in the Uy-
gur population is similar to those of other populations. To
our knowledge, no large-scale epidemiological study exam-
ining the relationship between serum UA and the prevalence
or severity of these disorders in Uygurs has been performed
to date. The present study aimed to clarify the relationship
between serum UA and these cardiovascular risk factors in
the Uygur population.
Materials and methods
Subjects
Two thousand two hundred ninety Uygur subjects, with
no intermarriages within the past 3 generations, were ran-
1033
domly selected by multistage cluster sampling from the He-
tian area in the south of the Xinjiang Uygur Autonomous
Region of China. Subjects with secondary hypertension,
acute stroke, gout, or cancer, or those taking diuretics, were
excluded from this study. After exclusion, 2127 subjects,
consisting of 859 males and 1268 females, aged 20 to
70 years, took part in the survey during the 2-month period
from January to March 2007, with an overall response rate
of 92.88%. Informed consent was obtained from all subjects
before any data collection or measurements. Hypertension
was diagnosed as systolic blood pressure
(SBP) ‡140 mm Hg and (or) diastolic blood pressure
(DBP) ‡90 mm Hg or antihypertension treatment according
to the World Health Organization (WHO) definition of hy-
pertension in 1999. DM was defined by a fasting blood glu-
cose (FBG) ‡7.0 mmol
L–1 (126 mg
dL–1) or a 2-h
postprandial glucose (2HPG) ‡11.1 mmol
L–1 (200 mg
dL–1).
FBG levels from 6.1 mmol
L–1 (110 mg
dL–1) to
6.9 mmol
L–1 (125 mg
dL–1) were defined as IFG. Two-
hour postprandial glucose levels from 7.8 mmol
L–1
(140 mg
dL–1) to 11.0 mmol
L–1 (199 mg
dL–1) were de-
fined as IGT. DM, IFG, and IGT were evaluated in 1761
participants who received oral glucose tolerance tests
(OGTT) (366 participants rejected OGTT). The homeosta-
sis model assessment insulin resistance index (HOMA-IR)
was used to assess IR, and IR was defined as HOMA-
IR > 1.75. Overweight and obesity were defined according
to the criteria of WHO (normal weight: body mass index
(BMI) < 25 kg
m–2; overweight: 25 kg
m–2 £ BMI £
30 kg
m–2; obesity: BMI > 30 kg
m–2). Dyslipidemia was
diagnosed according to the Guidelines for Prevention and
Control of Dyslipidemia in Chinese Adults: hypercholester-
emia (total cholesterol (TC) ‡ 6.22 mmol
L–1), hypo-HDL-c
(HDL-c < 1.04 mmol
L–1), hyper-low-density lipoprotein-
cholesterol (LDL-c) (LDC-c ‡ 4.14 mmol
L–1), and hyper-
triglyceridemia (triglyceride (TG) ‡ 2.26 mmol
L–1).
Hyperuricemia was diagnosed when serum UA was
>420 mmol
L–1 for men and >360 mmol
L–1 for women. The
risk index measurement for an individual was assessed by
how many cardiovascular risk factors he or she suffers
from, including hypertension, overweight–obesity, hy-
percholesteremia, hypertriglyceridemia, hyper-LDL-c,
hypo-HDL-c, hyperglycemia, IR, and smoking.
For example, the risk index was 3 if an individual suffered
from hypertension, obesity, and smoking. After categorizing
the subjects by the quartile of UA (UA1: UA £ 177 mmol
L–1;
UA2: 177 mmol
L–1 < UA £ 219 mmol
L–1; UA3:
219 mmol
L–1 < UA £ 271 mmol
L–1; UA4: UA >
271 mmol
L–1.), the odds ratios (ORs) of each cardiovascu-
lar risk factor were compared among different UA-based
groups. This study was approved by the Ethics Committee
of the People’s Hospital of Xinjiang Uygur Autonomous
Region (Xinjiang, China).
Phenotypic, demographic, and lifestyle data
A set of questionnaires was completed, including demo-
Published by NRC Research Press
 1034 Appl. Physiol. Nutr. Metab. Vol. 34, 2009

Table 1. The baseline characteristics compared by different sex and UA-based groups.

Male Female
Normouricemia, Hyperuricemia, Normouricemia, Hyperuricemia,
Characteristics n = 817 n = 42 p value n = 1247 n = 21 p value
Age (y) 56.79±13.9 50.95±12.6 0.008* 50.64±12.6 52.33±12.1 0.539
BMI (kg
m–2) 26.43±4.4 29.89±4.7 <0.001* 26.66±4.7 29.14±4.8 0.019*
SBP (mm Hg) 132.10±28.9 134.90±31.1 0.543 131.47±27.7 146.05±29.1 0.017*
DBP (mm Hg) 78.64±19.1 80.93±16.2 0.447 78.97±15.6 89.48±22.1 0.002*
BUN (mmol
L–1)
Appl. Physiol. Nutr. Metab. Downloaded from www.nrcresearchpress.com by TEXAS STATE UNIV on 08/09/13

5.88±1.7 6.92±1.8 <0.001* 5.05±1.6 7.16±2.8 <0.001*

Cr (mmol
L–1) 65.41±14.9 87.14±19.2 <0.001* 47.42±10.9 82.90±67.3 <0.001*
TC (mmol
L–1) 4.37±1.3 5.83±1.4 <0.001* 4.36±1.2 5.47±1.4 <0.001*
HDL-c (mmol
L–1) 1.05±0.4 1.23±0.4 0.001* 1.11±0.4 1.35±0.4 0.003*
LDL-c (mmol
L–1) 2.64±1.2 3.54±1.3 <0.001* 2.40±1.0 3.31±1.5 <0.001*
TG (mmol
L–1) 1.54±1.2 2.64±1.6 <0.001* 1.48±1.0 1.85±0.7 0.108
FBG (mmol
L–1) 5.81±2.7 7.12±2.3 0.002* 5.57±2.0 6.72±1.5 0.009*
2HPG (mmol
L–1) 7.98±5.3 8.03±4.4 0.964 7.98±4.5 9.29±3.8 0.252
LnFINS 2.5±0.3 2.96±0.4 0.024* 2.87±0.3 2.98±0.2 0.102
Ln3HPINS 3.14±0.3 3.16±0.3 0.553 3.27±0.3 3.43±0.3 0.047*
LnHOMA-IR 2.24±0.4 2.44±0.4 <0.001* 2.24±0.3 2.44±0.2 0.009*
Education (literacy %) 80.4 64.3 0.011* 91.6 81.0 0.085
Smoking (%) 27.8 57.1 <0.001* 1.0 4.8 0.087
Drinking (%) 13.6 35.7 <0.001* 0.1 0.6 0.731
Risk index 2.99±1.7 5.02±1.6 <0.001* 2.70±1.5 4.29±1.4 <0.001*
Note: Values are means ± standard deviation or percentage The differences between hyperuricemia and normouricemia were performed by
t test or by c2 test. BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; BUN, blood urea nitrogen; Cr, creati-
For personal use only.

nine; TC, serum total cholesterol; HDL-c, high-density lipoprotein cholesterol; LDL-c, low-density lipoprotein cholesterol; TG, triglyceride;
FBG, fasting blood glucose; 2HPG, 2-h postprandial glucose; FINS, fasting insulin; 3HPINS, 3-h postprandial insulin; HOMA-IR, homeostasis
model assessment insulin resistance index.
*Significant difference, p < 0.05.

graphic information; detailed history; family history of hy-
pertension, overweight–obesity, DM, and stroke; drug treat-
ment; education; alcohol consumption; and cigarette
smoking. Cigarette smoking was defined as having smoked
at least 1 cigarette per day for 1 or more years during the
participant’s lifetime. Alcohol use was assessed using an in-
terviewer-administered questionnaire with 6 questions re-
garding the frequency, amount, and type of alcoholic drinks
consumed. Participants were initially asked whether or not
they consumed alcohol. Participants who drank alcohol
were then asked to specify the number of years during
which they drank alcohol and the amounts of alcohol con-
sumption per month for each of 4 types of alcohol (beer,
liquor, wine (other than rice wine), and rice wine) over the
previous year. This quantity was then multiplied by the per-
centage of alcohol in each type of drink: for beer, 3.9%; for
wine, 11.6%; for rice wine, 13.6%; and for liquor, 53.3%.
Grams of alcohol per month were then summed across the
4 types of beverages and divided by 12.5 g to provide the
number of standardized alcoholic drinks consumed per
month. Alcohol consumption was defined as drinking at
least 12 drinks containing alcohol during the last year. Fol-
lowing a common protocol recommended by the American
Heart Association anthropometric measurements, the blood
pressure (BP) measurement was performed by trained and
certified observers 3 times per subject and the mean value
was considered the final BP value. Other data, such as
height, mass, and waistline and hip circumferences, were ob-
tained by standard protocols, and BMI was calculated. The
overnight fasting (12 h at least) venous blood was drawn
from all participants and 1761 participants received OGTT.
Serum was separated immediately and stored at –80 8C. All
blood samples were examined within a month in the Clinical
Center of the People’s Hospital of Xinjiang Uygur Autono-
mous Region.
Biological measurements
Serum lipids (including TC, TG, HDL-c, and LDL-c), se-
rum UA, FBG, 2HPG, blood urea nitrogen (BUN), creati-
nine (Cr), and other biological measurements were obtained
by enzymatic methods with an autoanalyzer (7600–010 Au-
tomatic Analyzer; HITACHI Medical System, Suzhou,
China). Fasting insulin (FINS) and 3-h postprandial insulin
(3HPINS) were determined by radioimmunity methods.
Quality controls were conducted by special docimaster.
Statistical analysis
Data analyses were performed using the SPSS statistics
package, version 15.0 (SPSS Inc., Chicago, Ill.). The catego-
rical variables were tested by c2 test. Quantitive variables
were shown as mean ± standard deviation and the differen-
ces among groups were evaluated by Student’s t test. FINS,
3HPINS, and HOMA-IR were transformed into normal dis-
tribution by function of base-e logarithm of nume  100.
(We multiplied FINS, 3HPINS, and HOMA-IR by 100 in or-
der to avoid a negative number after transform.) After ad-
justment for age, sex, education, smoking, and drinking,
logistic regression analysis was performed to assess the
UA-based risk for hypertension, overweight–obesity, dyslipi-
demia, hyperglycemia, and IR, respectively. The adjusted
factors also included lipid profiles, BMI, BUN, Cr, FBG,
2HPBG, FINS, and 3HPINS for hypertension; lipid profiles,

Published by NRC Research Press

 Appl. Physiol. Nutr. Metab. Downloaded from www.nrcresearchpress.com by TEXAS STATE UNIV on 08/09/13
For personal use only.

Li et al.

Table 2. The odds ratios (ORs) of each cardiovascular risk factor according to UA-based categories.

Factors UA1 (UA £ 177 mmol
L–1) UA2 (177 < UA £ 219 mmol
L–1) UA3 (219 < UA £ 271 mmol
L–1) UA4 (UA > 271 mmol
L–1) p value
Hypertension
OR 1 1.263 0.942 0.992 0.297
95% CI 0.905–1.764 0.659–1.347 0.647–1.520
Logistic regression model Logit (hypertension) = –6.763 + 0.046 (age) + 0.105 (BMI) + 0.213 (TG) + 0.212 (smoking) + 0.143 (drinking)
Overweight–obesity
OR 1 1.561 1.41 3.015 <0.001*
95% CI 1.114–2.187 0.999–1.991 1.996–4.555
Logistic regression model Logit (overweight–obesity) = –2.061 + 0.022 (age) – 0.728 (HDL-c) + 0.217 (LDL-c) + 0.481 (TG) + 1.104 (UA)
Hypercholesteremia
OR 1 1.973 2.715 5.14 <0.001*
95% CI 0.856–4.549 1.209–6.099 2.259–11.693
Logistic regression model Logit (hypercholesteremia) = –9.385 + 1.254 (female) + 0.233 (FBG) + 0.042 (Cr) + 1.637 (UA)
Hypo-HDL-c
OR 1 0.636 0.435 0.364 <0.001*
95% CI 0.458–0.881 0.305–0.619 0.238–0.555
Logistic regression model Logit (hypo-HDL-c) = 4.956 – 1.404 (female) – 0.239 (FBG) + 0.038 (BMI) – 1.011 (UA)
Hyper-LDL-c
OR 1 1.263 1.609 2.769 0.012*
95% CI 0.618–2.583 0.812–3.188 1.371–5.589
Logistic regression model Logit (hyper-LDL-c) = –5.184 + 0.018 (Cr) + 0.237 (FBG) + 1.018 (UA)
Hypertriglyceridemia
OR 1 2.387 3.528 6.236 <0.001*
95% CI 1.352–4.212 2.034–6.120 3.585–10.849
Logistic regression model Logit (hypertriglyceridemia) = –6.812 – 0.119 (BUN) + 0.161 (FBG) + 0.125 (BMI) + 1.830 (UA)
IR
OR 1 1.615 1.961 2.23 <0.001*
95% CI 1.233–2.116 1.474–2.608 1.604–3.100
Logistic regression model Logit (IR) = –5.680 + 0.453 (female) + 0.091 (BUN) + 0.169 (TC) + 0.093 (BMI) + 0.460 (TG) + 0.802 (UA)
Hyperglycemia
OR 1 0.853 0.913 1.199 0.196
95% CI 0.626–1.163 0.658–1.268 0.809–1.778
Logistic regression model Logit (hyperglycemia) = –6.044 + 0.524 (female) + 0.021 (age) + 0.015 (Cr) + 0.042 (BMI) + 0.445 (TC) + 0.043 (drinking)
Note: Logistic regression analysis was performed to determine ORs of each cardiovascular risk factor. The differences of ORs among various UA groups were compared with the lowest UA group. UA,
serum uric acid; CI, confidence interval; BMI, body mass index; TG, triglyceride; HDL-c, high-density lipoprotein cholesterol; LDL-c, low-density lipoprotein cholesterol; FBG, fasting blood glucose; Cr,
creatinine; BUN, blood urea nitrogen; IR, insulin resistance; TC, serum total cholesterol.
*The significant differences of ORs among UA1, UA2, UA3, and UA4 groups, p < 0.05.

Published by NRC Research Press

1035
 1036 Appl. Physiol. Nutr. Metab. Vol. 34, 2009

Table 3. Comparison of serum uric acid (UA) levels between population. The baseline characteristics of the study popula-
normal and disease groups by covariate variance analysis ad- tion are presented in Table 1, which shows that in the hyper-
justing for confounding factors. uricemia group, the mean values of BMI, BUN, Cr, TC,
HDL-c LDL-c, TG, FBG, LnFINS, LnHOMA-IR, and risk
Factors UA (mmol
L–1) f value p value
index significantly increased in males and BMI, SBP, DBP,
Hypertension BUN, Cr, TC, HDL-c LDL-c, FBG, LnHOMA-IR, and risk
Yes (n = 804) 239.11±79.112 0.184 0.668 index significantly increased in females. Compared with the
No (n = 1323) 223.64±73.574 hyperuricemia group, the subjects in the normouricemia
Overweight–obesity group were more highly educated and less likely to smoke
Appl. Physiol. Nutr. Metab. Downloaded from www.nrcresearchpress.com by TEXAS STATE UNIV on 08/09/13

Yes (n = 1315) 241.70±79.679 26.090 <0.001 or drink.

No (n = 812) 209.72±65.186 Adjusted for confounding factors, logistic regression anal-
Hypercholesteremia
ysis confirmed that the ORs for IR, overweight–obesity, hy-
Yes (n = 163) 295.23±91.032 24.077 <0.001
percholesteremia, hyper-LDL-c, and hypertriglyceridemia
No (n = 1964) 224.03±72.070
against the lowest UA group increased significantly with
UA, whereas it decreased with UA for hypo-HDL-c
Hypo-HDL-c (Table 2).
Yes (n = 1083) 217.56±72.610 22.038 <0.001 After adjusting for confounding factors, covariate var-
No (n = 1044) 241.86±77.614 iance analysis showed that the serum UA was significantly
Hyper-LDL-c higher in subjects with overweight–obesity, hyperglycemia,
Yes (n = 177) 280.42±90.627 12.580 <0.001 IR, hypercholesteremia, hypertriglyceridemia, and hyper-
No (n = 1950) 224.86±72.891 LDL-c. However, the serum UA in hypo-HDL-c subjects
Hypertriglyceridemia was significantly lower than that of normo-HDL-c subjects
Yes (n = 324) 277.01±88.575 50.486 <0.001 (Table 3).
No (n = 1803) 220.95±70.295 The c2 test indicated that the prevalence of overweight–
obesity, hypercholesteremia, hyper-LDL-c, hypertriglyceri-
IR
demia, hyperglycemia, and IR in hyperuricemic subjects
For personal use only.

Yes (n = 1027) 248.46±78.418 17.845 <0.001

No (n = 1100) 211.78±69.299
Hyperglycemia
Yes (n = 825) 248.36±84.686 8.292 0.004
No (n = 936) 216.62±66.848
Note: A value of p < 0.05 was considered significant. HDL-c,
high-density lipoprotein cholesterol; LDL-c, low-density lipoprotein
cholesterol; IR, insulin resistance.
SBP, DBP, BUN, Cr, FBG, 2HPBG, FINS, and 3HPINS for
overweight–obesity; lipid profiles, SBP, DBP, BUN, Cr, and
BMI for hyperglycemia; lipid profiles, SBP, DBP, BUN, Cr,
and BMI for IR; SBP, DBP, BUN, Cr, BMI, FBG, 2HPBG,
FINS, and 3HPINS for hypercholesteremia; SBP, DBP,
BUN, Cr, BMI, FBG, 2HPBG, FINS, and 3HPINS for hy-
pertriglyceridemia; SBP, DBP, BUN, Cr, BMI, FBG,
2HPBG, FINS, and 3HPINS for hyper-LDL-c; and SBP,
DBP, BUN, Cr, BMI, FBG, 2HPBG, FINS, and 3HPINS
for hypo-HDL-c. After adjusting for the same confounding
factors as in logistic regression analysis, covariate variance
analysis was performed to evaluate the difference of serum
UA levels between hypertension and normotension, over-
weight–obesity and normoweight, hyperglycemia and eugly-
cemia, hypercholesteremia and normocholesteremia,
hypertriglyceridemia and normotriglyceridemia, hyper-LDL-c
and normo-LDL-c, and hypo-HDL-c and normo-HDL-c, re-
spectively. The c2 test was used to compare the prevalence
of hypertension, overweight–obesity, dyslipidemia, hyper-
glycemia, and IR between hyperuricemia and normourice-
mia subjects. A p value less than 0.05 was statistically
significant.
Results
The prevalence of hyperuricemia was 3.1% in the Uygur
was significantly higher than in normouricemic subjects.
However, the prevalence of hypo-HDL-c in hyperuricemic
subjects was significantly lower than in normouricemic sub-
jects (Table 4).
Discussion
This cross-sectional study demonstrates that in Uygurs,
elevation of UA is associated with overweight–obesity,
hypercholesteremia, hyper-LDL-c, hypertriglyceridemia, hy-
perglycemia, and IR, which are metabolic risk factors for
CVD. Importantly, UA is positively correlated to HDL-c.
This is inconsistent with the previous opinion that UA is re-
lated negatively to HDL-c.
UA is the final oxidation product of purine catabolism.
Epidemiological, experimental, and clinical studies have
confirmed that serum UA is a predictor for overweight–obe-
sity, hypercholesteremia, hyper-LDL-c, hypertriglyceride-
mia, hyperglycemia, and IR (Nakagawa et al. 2006;
Krishnan et al. 2007; Lohsoonthorn et al. 2006; Choi and
Ford 2007; Hikita et al. 2007; Ford et al. 2007; Chen et al.
2008; Wang et al. 2007; Chien et al. 2008; Dehghan et al.
2008). Our results support this conclusion. After adjusting
for confounding factors (considering the substantial sex dif-
ferences in UA levels and these risk factors, sex was consid-
ered as a confounding factor), we have demonstrated that
elevated serum UA is associated with these cardiovascular
risk factors in Xinjiang Uygurs. The mechanisms underlying
the relationships between serum UA and these cardiovascu-
lar risk factors are unclear. It has been hypothesized that hy-
peruricemia and overweight–obesity, dyslipidemia, and
hyperglycemia are associated because they are components
of the metabolic syndrome (MS). IR or hyperinsulinemia is
possibly a common pathway that unites the various abnor-
malities of MS. Several studies have confirmed that under-

Published by NRC Research Press

 Li et al. 1037

Table 4. Comparison of the prevalence rates of hypertension, over-

weight–obesity, hypercholesteremia, hypo-HDL-c, hyper-LDL-c, hypertri-
glyceridemia, hyperglycemia, and insulin resistance (IR) between
hyperuricemia and normal groups.

Normal UA Hyperuricemia
Factors group group c2 p value
Hypertension
Yes (n = 804) 774 30 2.662 0.103
Appl. Physiol. Nutr. Metab. Downloaded from www.nrcresearchpress.com by TEXAS STATE UNIV on 08/09/13

No (n = 1323) 1290 33
Overweight–obesity
Yes (n = 1315) 1261 54 15.700 <0.001
No (n = 812) 803 9
Hypercholesteremia
Yes (n = 163) 140 23 76.336 <0.001
No (n = 1964) 1924 40
Hypo-HDL-c
Yes (n = 1083) 1062 21 8.032 0.005
No (n = 1044) 1002 42
Hyper-LDL-c
Yes (n = 177) 159 18 34.895 <0.001
No (n = 1950) 1905 45
Hypertriglyceridemia
Yes (n = 324) 298 26 34.086 <0.001
No (n = 1803) 1766 37
For personal use only.

IR
Yes (n = 1027) 980 47 18.010 <0.001
No (n = 1100) 1084 16
Hyperglycemia
Yes (n = 825) 776 49 34.114 <0.001
No (n = 936) 927 9
Note: A value of p < 0.05 was considered significant. UA, serum uric acid;
HDL-c, high-density lipoprotein cholesterol; LDL-c, low-density lipoprotein cho-
lesterol.

excretion of UA into the urine caused by the effect of in-
sulin on the urinary tubular tract has been demonstrated
with physiological hyperinsulinemia acutely reducing uri-
nary UA (Facchini et al. 1991; Quiñones Galvan et al.
1995; Ter Maaten et al. 1997); accelerated UA production
coupled with the synthesis of TG (Fabregat et al. 1987);
increased insulin secretion and resistance were observed in
obesity (Pi-Sunyer 2002); and UA is a surrogate of IR
(Chen et al. 2008). However, Lohsoonthorn et al. (2006)
noted that obesity and central body fat distribution, rather
than hyperinsulinemia–IR, play a major role in linking hy-
peruricemia with cardiovascular risk factor clustering in
MS. Some groups have found that the contribution of UA
as an additional component of MS seems to be insignifi-
cant (Liou et al. 2006). Therefore, the above-mentioned
hypothesis remains uncertain but merits further study.
Uygurs, with a total population of 8.8 million (45% of the
total population of Xinjiang, national census 2003), are the
largest nationality in Xinjiang. The present Uygur popula-
tion dwells in South Xinjiang, which is farthest from the
oceans in China. Therefore, they have little access to sea-
food. Uygurs have excessive starch intake because the popu-
lar and main food for them is ‘‘nang’’, which is made of
flour. They seldom have Western food and soft drinks,
which are sweetened by sugar or fructose. Fructose, rather
than starch, is the only sugar that can raise serum UA con-
centrations, and this has been shown in both humans (Stirpe
et al. 1970) and rodents (Stavric et al. 1976). Their diet may
be the reason for the lower prevalence of hyperuricemia in
Uygurs. Although the prevalence of hyperuricemia is very
low (3.1%), we found higher levels of serum UA to be asso-
ciated with an increased risk of overweight–obesity,
hypercholesteremia, hyper-LDL-c, hypertriglyceridemia,
hyperglycemia, and IR in the present Uygur population.
Moreover, the risk index of the cluster of hypertension,
overweight–obesity, dyslipidemia, hyperglycemia, and IR in-
creased with serum UA in this population, which indicates
that the possibility of these risk factors occurring together
increases with the serum UA levels. Previous studies have
confirmed that higher morbidity and mortality from CVD
are observed when these risk factors occur together. All
these data suggest a potential causative role for UA in the
development of cardiovascular risk. This hypothesis merits
investigation, perhaps including intervention studies to re-
duce UA. Moreover, even at the low levels commonly en-
countered in the Uygur population, serum UA appears to be
associated with established cardiovascular risk factors. Pro-
spective studies to evaluate the impact of serum UA on

Published by NRC Research Press

 1038
CVD risk in this ethnic group provide an opportunity to fur-
ther evaluate a causative role for serum UA in its pathogen-
esis. The ideal reference threshold of serum UA for efficient
prevention of CVD in the Uygur general population is un-
clear. Future large-scale prospective studies are needed to
clarify this point.
The relationship between hyperuricemia and hypertension
has been well reported in different ethnic groups, and hyper-
uricemia has been proposed as the key linking factor for hy-
Appl. Physiol. Nutr. Metab. Downloaded from www.nrcresearchpress.com by TEXAS STATE UNIV on 08/09/13
pertension (Sundström et al. 2005; Shankar et al. 2006;
Mellen et al. 2006; Forman et al. 2007; Perlstein et al.
2006). Furthermore, 2 groups have found that the lowering
of UA concentrations in patients with hyperuricemia results
in a significant fall in BP (Feig et al. 2004; Siu et al. 2006).
However, inconsistent with previous studies, our study
found only a nonsignificant trend for an association between
UA and hypertension in the present population. Though the
underlying reasons are still unclear, obesity, aging, Uygur-
specific environments, and lifestyle factors such as excessive
salt intake, less physical exercise, mental stress, and genetic
susceptibility should be considered. Further studies between
hypertension and ethnic-specific environment, lifestyles, and
genetic susceptibility are needed to clarify this point.
The other important finding of the present study is that
the HDL-c levels increased significantly with UA, whereas
the prevalence of hypo-HDL-c decreased with UA; this is
For personal use only.
inconsistent with previous studies (Hikita et al. 2007; Lin et
al. 2007). Also, Yan et al. (2005) discovered that the serum
HDL-c levels in a rural Uygur population in China increased
with age, BMI, BP, and fasting glucose levels. The underly-
ing causes for these novel phenomena remain unclear. In ad-
dition to the influence of ethnic-specific environmental and
behavioral factors, genetic factors might play a role. There-
fore, further study is needed to elucidate both the environ-
mental and genetic determinants of the novel HDL-c
distribution and their association with CVD in the Uygur
population. Furthermore, the role of an elevated HDL-c
level with regard to CVD risk in Uygurs is unclear and de-
serves further study as well.
The present study has several limitations. Because we
used cross-sectional data to predict the influences of serum
UA on the prevalence and severity of overweight–obesity,
dyslipidemia, hyperglycemia, and IR, a causal relationship
cannot be determined. In addition, our study subjects are
only from the Uygur population; though the homogenous
nature of the cohort eliminates confounding factors such as
race and socioeconomic status, our results may not necessa-
rily be generalizable to other populations.
In summary, the prevalence of overweight–obesity, hyper-
cholesteremia, hyper-LDL-c, hypertriglyceridemia, hyper-
glycemia, and IR increases with serum UA, although the
prevalence of hyperuricemia is lower in Uygurs. A notice-
able result of the present study is that the HDL-c signifi-
cantly increases but the prevalence of hypo-HDL-c
decreases with UA. Further research should explore the eth-
nic-specific genetic and environmental factors underlying
the lower prevalence of hyperuricemia and the positive asso-
ciation between serum UA and HDL-c in Uygurs.
Acknowledgements
We would like to express our sincere gratitude to
Appl. Physiol. Nutr. Metab. Vol. 34, 2009
M. Kuerban for his continuous support of our population
survey in the Hetian area. In addition, we thank all the staff
of the Center of Diagnosis, Treatment and Research of Hy-
pertension in Xinjiang for their support with the medical ex-
amination and demographic data collection. This study was
supported by the National Natural Science Foundation of
China (30260038) and the National Supporting Programs
for Critical Illness of China (2002BA711A0B).
References
Chen, L.K., Lin, M.H., Lai, H.Y., Hwang, S.J., and Chiou, S.T.
2008. Uric acid: a surrogate of insulin resistance in older wo-
men. Maturitas, 59(1): 55–61. doi:10.1016/j.maturitas.2007.10.
006. PMID:18082342.
Chien, K.L., Chen, M.F., Hsu, H.C., Chang, W.T., Su, T.C., Lee,
Y.T., and Hu, F.B. 2008. Plasma uric acid and the risk of type
2 diabetes in a Chinese community. Clin. Chem. 54(2): 310–
316. doi:10.1373/clinchem.2007.095190. PMID:18089655.
Choi, H.K., and Ford, E.S. 2007. Prevalence of the metabolic syn-
drome in individuals with hyperuricemia. Am. J. Med. 120(5):
442–447. doi:10.1016/j.amjmed.2006.06.040. PMID:17466656.
Conen, D., Wietlisbach, V., Bovet, P., Shamlaye, C., Riesen, W.,
Paccaud, F., and Burnier, M. 2004. Prevalence of hyperuricemia
and relation of serum uric acid with cardiovascular risk factors
in a developing country. BMC Public Health, 25: 4–9. PMID:
15043756.
Cooper, R.S., Orduñez, P., Iraola Ferrer, M.D., Munoz, J.L., and
Espinosa-Brito, A. 2006. Cardiovascular disease and associated
risk factors in Cuba: prospects for prevention and control. Am.
J. Public Health, 96(1): 94–101. doi:10.2105/AJPH.2004.
051417. PMID:16317211.
Dehghan, A., van Hoek, M., Sijbrands, E.J., Hofman, A., and
Witteman, J.C. 2008. High serum uric acid as a novel risk factor
for type 2 diabetes. Diabetes Care, 31(2): 361–362. doi:10.2337/
dc07-1276. PMID:17977935.
Fabregat, I., Revilla, E., and Machado, A. 1987. Short-term control
of the pentose phosphate cycle by insulin could be modulated by
NADPH/NADP ratio in rat adipocytes and hepatocytes. Bio-
chem. Biophys. Res. Commun. 146(2): 920–925. doi:10.1016/
0006-291X(87)90618-8. PMID:3304289.
Facchini, F., Chen, Y.D.I., Hollenbeck, C.B., and Reaven, G.M.
1991. Relationship between resistance to insulin-mediated glu-
cose uptake, urinary uric acid clearance, and plasma uric acid
concentration. JAMA, 266(21): 3008–3011. doi:10.1001/jama.
266.21.3008. PMID:1820474.
Feig, D.I., Nakagawa, T., Karumanchi, S.A., Oliver, W.J., Kang,
D.H., Finch, J., and Johnson, R.J. 2004. Hypothesis: Uric acid,
nephron number, and the pathogenesis of essential hypertension.
Kidney Int. 66(1): 281–287. doi:10.1111/j.1523-1755.2004.
00729.x. PMID:15200435.
Ford, E.S., Li, C., Cook, S., and Choi, H.K. 2007. Serum concentra-
tions of uric acid and the metabolic syndrome among US chil-
dren and adolescents. Circulation, 115(19): 2526–2532. doi:10.
1161/CIRCULATIONAHA.106.657627. PMID:17470699.
Forman, J.P., Choi, H., and Curhan, G.C. 2007. Plasma uric acid
level and risk for incident hypertension among men. J. Am.
Soc. Nephrol. 18(1): 287–292. doi:10.1681/ASN.2006080865.
PMID:17167112.
Guo, Y.Y., Wang, K., Zhao, L., and He, B.X. 2006. Epidemiology
study on metabolic syndrome in Xinjiang. Chin. J. Intern.. Med.
45: 227–228. [In Chinese.]
Hikita, M., Ohno, I., Mori, Y., Ichida, K., Yokose, T., and Hosoya,
T. 2007. Relationship between hyperuricemia and body fat dis-
Published by NRC Research Press
 Li et al.
tribution. Intern. Med. 46(17): 1353–1358. doi:10.2169/
internalmedicine.46.0045. PMID:17827832.
Krishnan, E., Kwoh, C.K., Schumacher, H.R., and Kuller, L. 2007.
Hyperuricemia and incidence of hypertension among men with-
out metabolic syndrome. Hypertension, 49(2): 298–303. doi:10.
1161/01.HYP.0000254480.64564.b6. PMID:17190877.
Li, Y., Stamler, J., Xiao, Z.H., Folsom, A., Tao, S.C., and Zhang,
H.Y.; The PRC-USA Collaborative Study in Cardiovascular and
Cardiopulmonary Epidemiology. 1997. Serum uric acid and its
correlates in Chinese adult populations, urban and rural, of
Appl. Physiol. Nutr. Metab. Downloaded from www.nrcresearchpress.com by TEXAS STATE UNIV on 08/09/13
Beijing. Int. J. Epidemiol. 26(2): 288–296. doi:10.1093/ije/26.2.
288. PMID:9169163.
Lin, J.D., Chiou, W.K., Chang, H.Y., Liu, F.H., and Weng, H.F.
2007. Serum uric acid and leptin levels in metabolic syndrome:
a quandary over the role of uric acid. Metabolism, 56(6): 751–
756. doi:10.1016/j.metabol.2007.01.006. PMID:17512306.
Liou, T.L., Lin, M.W., Hsiao, L.C., Tsai, T.T., Chan, W.L., Ho,
L.T., and Hwu, C.-M. 2006. Is hyperuricemia another facet of
the metabolic syndrome? J. Chin. Med. Assoc. 69(3): 104–109.
doi:10.1016/S1726-4901(09)70186-8. PMID:16599014.
Lohsoonthorn, V., Dhanamun, B., and Williams, M.A. 2006. Preva-
lence of hyperuricemia and its relationship with metabolic syn-
drome in Thai adults receiving annual health exams. Arch.
Med. Res. 37(7): 883–889. doi:10.1016/j.arcmed.2006.03.008.
PMID:16971230.
Mellen, P.B., Bleyer, A.J., Erlinger, T.P., Evans, G.W., Nieto, F.J.,
Wagenknecht, L.E., et al. 2006. Serum uric acid predicts inci-
dent hypertension in a biethnic cohort: the atherosclerosis risk
For personal use only.
in communities study. Hypertension, 48(6): 1037–1042. doi:10.
1161/01.HYP.0000249768.26560.66. PMID:17060502.
Nakagawa, T., Hu, H.B., Zharikov, S., Tuttle, K.R., Short, R.A.,
Glushakova, O., et al. 2006. A causal role for uric acid in fruc-
tose-induced metabolic syndrome. Am. J. Physiol. Renal Phy-
siol. 290(3): F625–F631. doi:10.1152/ajprenal.00140.2005.
PMID:16234313.
Nan, H., Qiao, Q., Dong, Y., Gao, W., Tang, B., Qian, R., and
Tuomilehto, J. 2006. The prevalence of hyperuricemia in a po-
pulation of the coastal city of Qingdao, China. J. Rheumatol.
33(7): 1346–1350. PMID:16821269.
Perlstein, T.S., Gumieniak, O.G., Williams, G.H., Sparrow, D.,
Vokonas, P.S., Gaziano, M., et al. 2006. Uric acid and the de-
velopment of hypertension: the normative aging study. Hyper-
tension, 48(6): 1031–1036. doi:10.1161/01.HYP.0000248752.
08807.4c. PMID:17060508.
Pi-Sunyer, F.X. 2002. The obesity epidemic: pathophysiology and
consequences of obesity. Obes. Res. 10(Suppl. 2): 97S–104S.
doi:10.1038/oby.2002.202. PMID:12490658.
Quiñones Galvan, A., Natali, A., Baldi, S., Frascerra, S., Sanna, G.,
Ciociaro, D., and Ferrannini, E. 1995. Effect of insulin on uric
acid excretion in humans. Am. J. Physiol. Endocrinol. Metab.
268(1 Pt. 1): E1–E5. PMID:7840165.
Shankar, A., Klein, R., Klein, B.E., and Nieto, F.J. 2006. The asso-
ciation between serum uric acid level and long-term incidence
of hypertension: population-based cohort study. J. Hum. Hyper-
tens. 20(12): 937–945. doi:10.1038/sj.jhh.1002095. PMID:
17024135.
1039
Shao, J.H., Mo, B.Q., Yu, R.B., Li, Z., Liu, H., and Xu, Y.C. 2003.
Epidemiological study on hyperuricemia and gout in community
of Nanjing. Chin J Dis Control Prev. 7: 305–308.
Shao, J.H., Shen, X., Li, D.Y., Shen, H.B., Xu, Y.C., and Mo, B.Q.
2007. Study on the relationship between compositions of hyper-
uricemia and metabolic syndrome. Zhonghua Liu Xing Bing
Xue Za Zhi, 28(2): 180–183. [In Chinese.] PMID:17649693.
Siu, Y.P., Leung, K.T., Tong, M.K., and Kwan, T.H. 2006. Use of
allopurinol in slowing the progression of renal disease through
its ability to lower serum uric acid level. Am. J. Kidney Dis.
47(1): 51–59. doi:10.1053/j.ajkd.2005.10.006. PMID:16377385.
Stavric, B., Johnson, W.J., Clayman, S., Gadd, R.E., and Chartrand,
A. 1976. Effect of fructose administration on serum urate levels
in the uricase inhibited rat. Experientia, 32(3): 373–374. doi:10.
1007/BF01940847. PMID:1253916.
Stirpe, F., Della Corte, E., Bonetti, E., Abbondanza, A., Abbati, A.,
and De Stefano, F. 1970. Fructose-induced hyperuricaemia. Lan-
cet, 2(7686): 1310–1311. doi:10.1016/S0140-6736(70)92269-5.
PMID:4098798.
Strasak, A.M., Kelleher, C.C., Brant, L.J., Rapp, K., Ruttmann, E.,
Concin, H., et al. 2008. Serum uric acid is an independent pre-
dictor for all major forms of cardiovascular death in 28,613 el-
derly women: a prospective 21-year follow-up study. Int. J.
Cardiol. 125(2): 232–239. doi:10.1016/j.ijcard.2007.11.094.
PMID:18237790.
Sun, Y.P., Yao, H., Amulajiang, M., and Cai, Z.H. 2007. The ana-
lysis of uric acid level and the prevalence of hyperuricmia in
Uygur ethnic population. J. Xinjiang Medical University, 5:
458–460. [In Chinese.]
Sundström, J., Sullivan, L., D’Agostino, R.B., Levy, D., Kannel,
W.B., and Vasan, R.S. 2005. Relations of serum uric acid to
longitudinal blood pressure tracking and hypertension incidence.
Hypertension, 45(1): 28–33. PMID:15569852.
Tanuseputro, P., Manuel, D.G., Leung, M., Nguyen, K., and
Johansen, H.; Canadian Cardiovascular Outcomes Research
Team. 2003. Risk factors for cardiovascular disease in Canada.
Can. J. Cardiol. 19(11): 1249–1259. PMID:14571310.
Ter Maaten, J.C., Voorburg, A., Heine, R.J., Ter Wee, P.M.,
Donker, A.J., and Gans, R.O. 1997. Renal handling of urate and
sodium during acute physiological hyperinsulinaemia in healthy
subjects. Clin. Sci. (Lond.), 92(1): 51–58. PMID:9038591.
Wang, M., Zhao, D., Wang, W., Liu, J., Liu, J., and Liu, S. 2007.
A prospective study on relationship between blood uric acid le-
vels, insulin sensitivity and insulin resistance. Zhonghua Nei Ke
Za Zhi, 46(10): 824–826. [In Chinese.] PMID:18218233.
Yan, W., Gu, D., Yang, X., Wu, J., Kang, L., and Zhang, L. 2005.
High-density lipoprotein cholesterol levels increase with age,
body mass index, blood pressure and fasting blood glucose in a
rural Uygur population in China. J. Hypertens. 23(11): 1985–
1989. doi:10.1097/01.hjh.0000187254.12375.b6. PMID:
16208139.
Zhu, S., Tang, P., Xie, L., Zhou, L.C., Zhang, J., Wang, J., et al.
2002. Epidemiological survey (1999–2000) on cardiovascular
risk factors in chengdu-hyperuric acid and clinical implication.
Chin. J. Hypertension, 5: 476–478. [In Chinese.]
Published by NRC Research Press
 This article has been cited by:

1. Ji-Rong Yue, Chang-Quan Huang, Bi-Rong Dong. 2012. Association of serum uric acid with body mass index among long-lived
Chinese. Experimental Gerontology 47:8, 595-600. [CrossRef]
2. Li Nanfang, Guo Yanying, Wang Hongmei, Yan Zhitao, Zhang Juhong, Zhou Ling, Luo Wenli. 2010. Variations of Six
Transmembrane Epithelial Antigen of Prostate 4 (STEAP4) Gene Are Associated with Metabolic Syndrome in a Female Uygur
General Population. Archives of Medical Research 41:6, 449-456. [CrossRef]
Appl. Physiol. Nutr. Metab. Downloaded from www.nrcresearchpress.com by TEXAS STATE UNIV on 08/09/13
For personal use only.