Antioxidant properties of

high-density lipoproteins are

impaired in women with polycystic
ovary syndrome
Jinxia Zhang, M.D.,a Yujin Zhang, M.D.,b Hongwei Liu, M.D.,b Huai Bai, M.D., Ph.D.,a Ying Wang, M.D.,b
Changan Jiang, Ph.D.,a and Ping Fan, M.D., Ph.D.a
a
Laboratory of Genetic Disease and Perinatal Medicine, Key Laboratory of Birth Defects and Related Diseases of Women
and Children, Ministry of Education, West China Second University Hospital, Sichuan University; and b Department of
Obstetrics and Gynecology, West China Second University Hospital, Sichuan University, Chengdu, Sichuan, People's
Republic of China

Objective: To determine the relationships among the inflammatory index, intrinsic oxidation levels, lipid and apolipoprotein (apo)A-I
concentrations of high-density lipoprotein (HDL), and polycystic ovary syndrome (PCOS).
Design: Cross-sectional study.
Setting: University hospital.
Patient(s): A total of 425 patients with PCOS and 441 control women were included.
Intervention(s): None.
Main Outcome Measure(s): The HDL inflammatory index (HII) was determined using a cell-free fluorometric assay. Intrinsic HDL oxidation
levels, HDL-free cholesterol, HDL-cholesterol ester, HDL-triglyceride, serum apoA-I, and malondialdehyde levels were also measured.
Result(s): The mean HII value and the frequency of HII R1 were significantly higher in the PCOS group (0.77
0.54, 27.1%) than in
the control group (0.53
0.37, 8.4%). These values were also higher in each of the 4 PCOS phenotypes based on the Rotterdam criteria
than in the controls, and higher in patients with hyperandrogenism (HA) þ oligo- and/or anovulation (OA) phenotype than in those
with OA þ polycystic ovary (PCO) phenotype. Furthermore, patients with PCOS with OA þ PCO had lower malondialdehyde and
intrinsic HDL oxidation levels compared with those with HA. Multivariate regression analysis demonstrated that PCOS, HDL-
cholesterol ester, and E2 levels were the main predictors of HII value.
Conclusion(s): The impairment of HDL antioxidant/anti-inflammatory function in PCOS is related to HA status, increased oxidative
stress, and abnormalities in HDL components and thus may contribute to PCOS pathogenesis
and increase the risks of future cardiovascular diseases. (Fertil SterilÒ 2015;103:1346–54.
Ó2015 by American Society for Reproductive Medicine.) Use your smartphone
Key Words: Polycystic ovary syndrome, dysfunctional high-density lipoprotein, oxidative to scan this QR code
stress, inflammation, dyslipidemia and connect to the
discussion forum for
this article now.*
Discuss: You can discuss this article with its authors and with other ASRM members at http://
fertstertforum.com/zhangj-antioxidant-properties-hdl-pcos/ * Download a free QR code scanner by searching for “QR
scanner” in your smartphone’s app store or app marketplace.

P
olycystic ovary syndrome (PCOS) affects 6%–15% of women of repro- long-term health risks, such as obesity
is a common female endocrine ductive age (1). Polycystic ovary (2), dyslipidemia (3), increased oxida-
and metabolic disorder that syndrome is often associated with tive stress (4, 5), chronic inflammation
(6), elevated metabolic syndrome (MS),
Received October 15, 2014; revised February 16, 2015; accepted February 17, 2015; published online
March 23, 2015. and type 2 diabetes risks (7). Several
J.Z. has nothing to disclose. Y.Z. has nothing to disclose. H.L. has nothing to disclose. H.B. has nothing studies have also demonstrated an
to disclose. Y.W. has nothing to disclose. C.J. has nothing to disclose. P.F. has nothing to disclose.
Financial support was received from the Chinese National Natural Science Foundation (grants increase in subclinical atherosclerosis
81070463 and 81370681) and the Program for Changjiang Scholars and Innovative Research in PCOS, as measured by carotid artery
Team in University, Ministry of Education (grant IRT0935).
J.Z. and Y.Z. should be considered similar in author order.
intima media thickness, coronary
Reprint requests: Ping Fan, M.D., Ph.D., Sichuan University, Chengdu, West China Second Hospital, artery calcium scores, and endothelial
Laboratory of Genetic Disease and Perinatal Medicine, Sichuan 610041, China (E-mail: dysfunction (8). A systematic review
fanping15@scu.edu.cn).
and meta-analysis has revealed a sig-
Fertility and Sterility® Vol. 103, No. 5, May 2015 0015-0282/$36.00 nificant twofold risk of coronary artery
Copyright ©2015 American Society for Reproductive Medicine, Published by Elsevier Inc.
http://dx.doi.org/10.1016/j.fertnstert.2015.02.024 disease and stroke for patients with

1346 VOL. 103 NO. 5 / MAY 2015


PCOS relative to women without PCOS, not fully depending
on body mass index (BMI) (9). However, few prospective
studies have examined nonfatal and fatal cardiac events in
women with well-defined PCOS (8, 9). The pathogenesis of
PCOS remains unclear, but studies have suggested that
PCOS has a complex, multifactorial etiology resulting from
the interactions between genetic, environmental, and
intrauterine factors (10).
Studies have indicated that atherosclerosis is a chronic
inflammatory disease of the arterial wall that is mediated in
part by oxidized low-density lipoproteins (LDLs) (11, 12).
High-density lipoproteins (HDLs) possess many important
atheroprotective functions. In addition to mediating reverse
cholesterol transport, they also have antioxidant, anti-
inflammatory, antithrombotic, and antiapoptotic activities
(11, 13). In mouse and human, the main component of HDL,
apolipoprotein (apo)A-I, is capable of removing ‘‘seeding
molecules’’ of lipid peroxidation from LDL, thus preventing
further oxidation of LDL-derived phospholipids and dramat-
ically reducing the inflammatory properties of LDL (11). High-
density lipoprotein–associated antioxidant enzymes, such as
paraoxonases (PON1, PON3) and platelet-activating factor
acetylhydrolase (PAF-AH), can prevent or destroy the forma-
tion of oxidized lipids (11). High-density lipoprotein also
seems to be the major carrier of lipid hydroperoxides in
plasma and may transport oxidized cholesteryl esters to the
liver for excretion (14, 15). Through these distinct
mechanisms, HDLs degrade and remove oxidized lipids and
decrease atherosclerosis risk (14). The antioxidant/anti-
inflammatory activities of HDLs are dependent on the pres-
ence of HDL-binding apo and antioxidant enzymes (11, 16).
Modification of these proteins, such as oxidation, may
weaken the antioxidant/anti-inflammatory activities of
HDLs and even convert them from anti-inflammatory to
proinflammatory particles (11).
The inflammatory/anti-inflammatory properties of HDL
could be evaluated by a cell-free assay measuring the HDL
inflammatory index (HII) (11, 17–19). The HII represents
the net action of all components in HDL, including
oxidized phospholipids, lipid hydroperoxides, HDL-
associated antioxidases and apolipoproteins, serum amyloid
A, and antioxidant vitamins (11). The impaired antioxidant/
anti-inflammatory activities of HDL have been associated
with acute coronary syndrome (20) and type 2 diabetes (17).
Mounting data have demonstrated that patients with
PCOS are associated with abnormal circulating markers of
oxidative stress, such as increased malondialdehyde (MDA),
homocysteine, asymmetric dimethylarginine, and superoxide
dismutase activity, as well as decreased glutathione levels and
PON1 activity (4). Our previous studies demonstrated that
serum apoA-I levels, HDL-associated PAF-AH (H-PAF-AH),
and apoE-containing H-PAF-AH activities were lower, and
LDL-associated PAF-AH (L-PAF-AH) activities as well as
the ratio of L-PAF-AH to H-PAF-AH activities were higher
in patients with PCOS (3, 5, 21). We observed that the PAF-
AH G994T gene mutation, which completely abolishes PAF-
AH activity, and the PON1 Q192R gene variation are the
risk factors for PCOS (22, 23). These results suggest that
impaired antioxidant activities of HDL may be associated
VOL. 103 NO. 5 / MAY 2015
Fertility and Sterility®
with increased oxidative stress in women with PCOS. In this
study, we investigated the relationships among the HII,
intrinsic oxidation levels, and lipid and apoA-I concentra-
tions of HDL and PCOS.
MATERIALS AND METHODS
Subjects
Women with or without PCOS, aged 20–40 years, were re-
cruited during 2006–2013 from the Outpatient Clinic of
Reproductive Endocrinology, West China Second University
Hospital, Sichuan University. All participants gave their
informed consent, and the study was approved by the institu-
tional review board of the West China Second University Hos-
pital, Sichuan University.
Polycystic ovary syndrome was defined by the presence
of two or more of the following features based on the
revised 2003 Rotterdam diagnostic criteria (24): oligo- or
anovulation (OA), which was assessed as oligomenorrhea
(fewer than eight cycles per year); biochemical and/or clin-
ical hyperandrogenism (HA), which was assessed by total T
(TT) levels above the 95th percentile (2.60 nmol/L) of the
levels that were detected in a group of normal menstru-
ating women with normal cycles (22); clinical presence of
obvious acne, which was defined as the number of come-
dones (>10) and inflammatory papules/pustules, as well
as nodules/cysts (>10) spread on the face, back, and chest
(3, 25, 26) and/or hirsutism with a modified Ferriman–
Gallwey (F-G) score of more than 6 (3, 27, 28); polycystic
ovaries (PCOs) were confirmed if there were 12 or more
follicles in each ovary measuring 2–9 mm in diameter
and/or increased ovarian volume (>10 mL) by ultrasonic
examination, with exclusion of other etiologies such as
congenital adrenal hyperplasia, androgen-secreting carci-
nomas, and Cushing's syndrome. All of the control women
had regular menstrual cycles (between 21 and 35 days), ex-
hibited normal circulating androgen levels, the absence of
obvious acne or hirsutism on physical examination, and
normal ovarian morphology as determined by ultrasound.
Subjects were excluded if they met one of the following
criteria: [1] clinically evident chronic or acute diseases, such
as infection, tumors, thyroid dysfunction, cardiovascular dis-
ease, endometriosis, hyperprolactinemia, hypogonadotropic
hypogonadism, or premature ovarian failure; [2] pregnant
or in the luteal phase according to P measurement
(>9.54 nmol/L); [3] taking medication known to affect the
metabolism of carbohydrates, lipids, or hormones within
3 months before the study; [4] smokers; [5] diabetes patients
with fasting glucose R7.0 mmol/L and/or 2-hour glucose
R11.1 mmol/L.
Clinical and anthropometric variables, including waist
circumference, hip circumference, waist-to-hip ratio, BMI
(kg/m2), systolic and diastolic blood pressure (SBP and
DBP), and the degree of hirsutism and acne, were evaluated
in all of the subjects. Ultrasound ovarian volume was also as-
sessed using the formula for the volume of an ellipsoid (29):
0.523  length  width  thickness.
Blood samples were obtained in the morning after an
overnight fast on the 3rd–10th days of the menstrual cycle
1347
ORIGINAL ARTICLE: REPRODUCTIVE ENDOCRINOLOGY
from regularly menstruating women or at random from
oligo-/amenorrheic women, placed on ice immediately, and
centrifuged at 1,500  g for 15 minutes at 4 C within 2 hours.
Serum and plasma samples were stored in 200–300mL ali-
quots at 80 C for later analysis. After baseline blood sam-
pling, an oral glucose tolerance test was immediately
performed: 75 g glucose was administered orally, and plasma
glucose and insulin levels were determined after 30, 60, 90,
120, and 180 minutes.
Measurements of HII and Intrinsic HDL Oxidation
Levels
The HDL inflammatory index measures the ability of
apoB-depleted serum to inhibit or enhance LDL oxidation
in the presence of a fluorescent substrate dichlorofluores-
cein (DCFH). The assay was performed essentially as previ-
ously described (17, 19, 20), with some modifications.
Briefly, standard LDL (density ¼ 1.019–1.063 g/mL) was
isolated from pooled fresh sera of healthy volunteers
by sequential ultracentrifugation (30). After dialyzing
against phosphate-buffered saline (PBS), a 50% sucrose so-
lution was added at a ratio of 1 volume of sucrose to 4 vol-
umes of LDL (19, 31). Cholesterol concentrations were
measured for sucrose-containing purified LDL before
it was divided into aliquots, and it was kept frozen
at 40 C until use. High-density lipoprotein was isolated
using the dextran sulfate (Mr 36000–50000; MP Biomedi-
cals) method (19, 32). Precipitate reagent (10% dextran
sulfate and 0.5 mol/l MgCl2; 40 mL) was mixed with
400 mL test serum containing 10% sucrose (50% sucrose
solution: serum ¼ 1:4), kept on ice for 10–20 minutes,
and subsequently centrifuged at 9,000  g for 20
minutes. The HDL-containing supernatant was used in
the experiments. Supernatant cholesterol (HDL-TC) was
quantified with an enzymatic assay (Siemens Healthcare
Diagnostics). Dichlorofluorescein diacetate (Sigma-Aldrich)
was first dissolved in fresh methanol at 2.0 mg/mL and
incubated at room temperature for 30 minutes in the dark
for DCFH release. To optimize the redox reaction, 25 mL
LDL (160 mg LDL cholesterol/mL in PBS) was added to a
black, flat-bottomed, 96-well polystyrene microtiter plate
(Thermo Scientific Nunc) and incubated at 37 C for 2 hours
with rotation. Next, 125 mL test HDL (64 mg HDL choles-
terol/mL in PBS) was mixed with LDL and incubated at
37 C on a rotator for 1 hour. Next, 25 mL DCFH solution
(0.2 mg/mL in PBS) was added to each well, including those
without lipoproteins, with LDL, LDLþHDL, or HDL, mixed,
and incubated at 37 C for 2 hours with rotation in the dark.
Fluorescence intensity was determined using the Varioskan
Flash Multimode Microplate Spectrophotometer (Thermo
Scientific) at an excitation wavelength of 509 nm, with a
bandwidth of 5 nm and an emission wavelength of
529 nm. The fluorescence values of wells without lipopro-
teins were subtracted from the values of the wells with
lipoproteins, whereas the LDL value in the absence of test
HDL was normalized to 1.0. The HII was calculated using
the following formula: [(LDLþHDL) value  HDL control
value]/LDL value. Test HDLs with HII >1 indicated
1348
pro-oxidation; those with HII <1 indicated anti-oxidation
(11, 17, 18).
Intrinsic HDL oxidation was measured using the fluores-
cence intensity resulting from the interaction of DCFH with
HDL alone. For each subject, 8 mg HDL in 150 mL PBS was
incubated with 25 mL DCFH solution (0.2 mg/mL) at 37 C
for 2 hours. Fluorescence values of the wells without HDL
were subtracted from the values of the wells with HDL. The
results were expressed as relative fluorescence units (rfu; 1
rfu ¼ 1,000 fluorescence intensity values) per mL undiluted
serum.
Samples were plated in quadruplicate and a pooled serum
sample from healthy volunteers was included on each plate as
a quality control. The intra- and interassay coefficients of
variation for all measurements were 3.4%–10.0% and
11.1%, respectively.
Analysis of MDA, apoA-I, and Hormonal and
Metabolic Markers
Serum FSH, LH, TT, E2, P, PRL, cortisol, and TSH levels, as well
as plasma insulin and glucose, total cholesterol (TC), HDL-
cholesterol (HDL-C), LDL-cholesterol (LDL-C), triglyceride
(TG), and apoA-I concentrations, the homeostasis model insu-
lin resistance index (HOMA-IR) and the atherogenic index
(AI) were measured or assessed as described previously (3).
The HDL-free cholesterol (HDL-FC) and HDL-TG were deter-
mined by enzymatic assay. Serum MDA levels were deter-
mined by spectrophotometry using micro-MDA kits
(NanJing Jiancheng Bioengineering Institute). The intra-
and interassay coefficients of variation for all measurements
were less than 5% and 10%, respectively.
ðHDL-CEÞ ¼ ðHDL-TCÞ  ðHDL-FCÞ:
Metabolic syndrome was defined as the presence of three
of the following five criteria: waist circumference R80 cm;
fasting TG R1.7 mmol/L; fasting HDL-C <1.29 mmol/L;
blood pressure R130/85 mm Hg; fasting glucose levels
R5.6 mmol/L (3).
Impaired glucose tolerance (IGT) was defined as a 2-hour
glucose result between 7.8 and 11.0 mmol/L.
Statistical Analyses
Results are presented as the mean
SD. Differences in variables
were evaluated by the independent-sample t test or nonpara-
metric tests (Mann–Whitney U test) between PCOS and control
subjects. Differences in percentages were evaluated by c2
tests between PCOS and control subjects or between subgroups.
Differences in variables between subgroups were analyzed by
analysis of variance (least significant difference and Dunnett's
T3) or analysis of covariance. Pearson correlation was per-
formed to define the correlations between HII or intrinsic HDL
oxidation levels and the other parameters in patients with
PCOS. Multivariate stepwise regression analyses were used to
assess the effect of other parameters, including PCOS, MS,
IGT, age, BMI, SBP, DBP, waist circumference, F-G score,
average ovarian volume, E2, TT, FSH, LH, fasting and 2-hour in-
sulin and glucose levels, TG, LDL-C, HDL-FC, HDL-CE, HDL-TG,
VOL. 103 NO. 5 / MAY 2015
 Fertility and Sterility®

apoA-I, and MDA levels on HII, and intrinsic HDL oxidation
levels in all of the subjects. We excluded the intrinsic HDL
oxidation level as an independent variable of the model when
HII was the dependent variable because it was in the HII calcu-
lation formula. We also excluded some parameters, such as
waist-to-hip ratio, LH/FSH ratio, HOMA-IR, TC, HDL-C, AI,
and TG/HDL-C ratio, in the regression models owing to the
presence of collinearity between them and some of the inde-
pendent variables. A P value of < .05 was considered to be
statistically significant. All of the statistical analyses were
performed using the Statistical Package for Social Sciences
13.0 for Windows (SPSS).
RESULTS
Selection of Subjects
Women with or without PCOS (n ¼ 1,689) were recruited in
the same manner during the same period. Of these subjects,
703 women met the revised 2003 Rotterdam diagnostic
criteria for PCOS, and 694 met the inclusion criteria for con-
trol women. After excluding subjects in the luteal phase by P
measurement (>9.54 nmol/L), those taking medication
known to affect carbohydrate and lipid metabolism or hor-
mone levels within 3 months before the study, smokers, and
those with diabetes, 425 patients with PCOS and 441 control
women were finally included in the present study.
The controls consisted of [1] infertile women (59.9%)
owing to fallopian obstruction (28.8%) or male factor infertility,
[2] women presenting for preconception counseling (35.8%),
and [3] healthy volunteers, doctors, and nurses (4.3%).
Clinical Characteristics, Hormonal Levels, and
Metabolic Profiles
As demonstrated in Table 1, BMI, waist circumference, waist-
to-hip ratio, diastolic BP, F-G score, and average ovarian
volume were significantly higher and the age was lower in
the PCOS group compared with the control group.
Because mean age and BMI were different between the
PCOS and the control groups, differences that could bias com-
parisons of hormonal levels and the metabolic profile between
the groups were adjusted for the difference in age and BMI.

TABLE 1

Clinical characteristics, hormonal levels, metabolic profiles, HII, and HDL-related components in patients with PCOS and control women.
Variable Controls (n [ 441) PCOS (n [ 425) P value P valuea
Age (y) 28.12
4.06 25.13
3.72 < .001
BMI (kg/m2) 21.00
2.73 22.85
4.18 < .001
Waist circumference (cm) 73.10
7.95 79.07
11.21 < .001
Waist-to-hip ratio 0.81
0.06 0.85
0.07 < .001
F-G score 0.20
0.66 1.78
2.02 < .001
SBP (mm Hg) 113.22
11.73 114.47
11.15 .110
DBP (mm Hg) 73.50
8.72 75.11
8.68 .007
Ovarian volume (mL) 7.78
2.69 10.28
4.25 < .001
Hormonal levels and metabolic profiles
E2 (pmol/L) 348.27
267.65 302.99
297.74 .056 .550
TT (nmol/L) 1.56
1.04 2.41
0.76 < .001 < .001
LH (IU/L) 8.62
11.40 14.09
11.92 < .001 < .001
FSH (IU/L) 6.59
2.96 5.98
2.37 .001 .046
LH/FSH 1.31
1.37 2.39
1.29 < .001 < .001
Fasting insulin (pmol/L) 65.28
35.49 97.49
63.01 < .001 < .001
2-h insulin (pmol/L)b 370.03
312.18 683.24
518.66 < .001 < .001
Fasting glucose (mmol/L) 5.30
0.46 5.34
0.41 .106 .700
2-h glucose (mmol/L)b 6.00
1.31 6.79
1.51 < .001 < .001
HOMA-IR 2.25
1.37 3.40
2.36 < .001 < .001
TG (mmol/L) 1.01
0.53 1.40
1.16 < .001 < .001
TC (mmol/L) 4.24
0.70 4.40
0.83 .002 < .001
HDL-C (mmol/L) 1.51
0.32 1.40
0.36 < .001 .094
LDL-C (mmol/L) 2.34
0.61 2.52
0.78 < .001 < .001
AI 1.90
0.66 2.32
0.99 < .001 < .001
TG/HDL-C 0.74
0.62 1.18
1.44 < .001 < .001
MS, % (n) 5.7 (25) 27.5 (117) < .001
IGT, % (n)b 9.2 (24) 22.5 (86) < .001
HII, HDL-related components and serum MDA levels
HII 0.53
0.37 0.77
0.54 < .001 < .001
HII R1, % (n) 8.4 (37) 27.1 (115) < .001
Intrinsic HDL oxidation (rfu/mL) 1.57
0.74 1.44
0.79 .015 .737
HDL-FC (mmol/L) 0.16
0.07 0.15
0.07 .001 .043
HDL-CE (mmol/L) 1.44
0.32 1.35
0.35 < .001 .295
HDL-TG (mmol/L) 0.27
0.14 0.29
0.15 .164 .121
ApoA-I (g/L) 1.43
0.19 1.38
0.20 .001 .516
MDA (nmol/mL) 3.60
1.21 4.24
1.60 < .001 < .001
Note: Values are presented as mean
SD unless otherwise noted.
a
All parameter comparisons were corrected for the differences in age and BMI between the two groups.
b
Control group (n ¼ 260), PCOS group (n ¼ 382).
Zhang. Antioxidant properties of HDL in PCOS. Fertil Steril 2015.

VOL. 103 NO. 5 / MAY 2015 1349

ORIGINAL ARTICLE: REPRODUCTIVE ENDOCRINOLOGY

Patients with PCOS had significantly higher TT and LH
levels, an LH-to-FSH ratio, fasting and 2-hour insulin con-
centrations, 2-hour glucose levels, HOMA-IR, TG, TC, LDL-C
levels, AIs, the TG/HDL-C ratio, MS, and IGT prevalence,
and they also had lower or tended to have decreased FSH
and HDL-C levels compared with the controls (Table 1).
HII Value, Prevalence of HII ‡1, MS, and IGT
As demonstrated in Table 1, the mean HII value, the preva-
lence of HII R1, MS, and IGT were significantly higher in
the PCOS group than in the control group.
We further analyzed the HII value, the frequencies of HII
R1, MS, and IGT in control women and in patients with
different key clinical PCOS phenotypes: [1] HA þ OA þ
PCO, [2] HA þ OA, [3] HA þ PCO, and [4] OA þ PCO
(Table 2). Compared with the control group, the HII values,
the HII R1 frequency, and MS prevalence were higher in
each of the four PCOS phenotypes (P< .05); IGT prevalence
was higher in the HA þ OA þ PCO subgroup and the OA þ
PCO subgroup (P< .01) and tended to be increased in the
HA þ OA subgroup (P¼ .054). The HII values and the HII
R1 frequency were also higher in the HA þ OA subgroup
than in the OA þ PCO subgroup (P< .05). Furthermore,
compared with the non-HA PCOS subgroup (OA þ PCO, n
¼ 139), the PCOS subgroup with HA (n ¼ 286) had higher
HII values (0.68
0.53 vs. 0.81
0.54, P¼ .019) and tended
to have increased HII R1 frequencies (21.6% vs. 29.7%,
P¼ .076) but decreased MS prevalence (36.7% vs. 23.1%,
P¼ .003).
HDL-Related Components and Serum MDA Levels
Patients with PCOS had significantly higher serum MDA
levels and lower HDL-FC levels than the controls (Table 1).
Compared with the control group, each of the four PCOS
phenotypes had increased MDA levels, patients with HA þ OA
þ PCO, HA þ OA, or OA þ PCO had decreased HDL-CE and
apoA-I levels, patients with HA þ OA þ PCO or OA þ PCO
had decreased HDL-FC levels, and patients with OA þ PCO
had decreased intrinsic HDL oxidation levels (Table 2).
Compared with the PCOS subgroup with HA (n ¼ 286), the
OA þ PCO subgroup (n ¼ 139) also had lower intrinsic HDL
oxidation levels, HDL-FC, HDL-CE, apoA-I, and MDA con-
centrations (all P< .05).
Relationship Between HII Values, Intrinsic HDL
Oxidation Levels, and Clinical and Metabolic
Parameters
Bivariate analysis including all of the continuous variables in
Table 1 in the patients with PCOS demonstrated that the HII
value correlated positively with intrinsic HDL oxidation
levels, HDL-CE, E2, HDL-FC, apoA-I, HDL-C, the F-G score,
and HDL-TG levels (r ¼ 0.457, 0.328, 0.229, 0.206, 0.206,
0.158, 0.126, and 0.104, respectively; all P< .05) and nega-
tively with waist-to-hip ratio, waist circumference, AI,
HOMA-IR, fasting insulin, and TG levels (r ¼ 0.201,
0.170, 0.121, 0.109, 0.108 and 0.104, respectively;
P< .05). Intrinsic HDL oxidation levels correlated positively
with HDL-CE, apoA-I, HDL-C, HDL-FC, E2, HDL-TG, and the
F-G score (r ¼ 0.664, 0.480, 0.422, 0.348, 0.149, 0.148, and
0.123, respectively; all P< .05) and negatively with AI, the
TG/HDL-C ratio, waist circumference, BMI, TG, waist-to-hip
ratio, fasting insulin levels, 2-hour insulin levels, HOMA-IR,
and FSH levels (r ¼ 0.289, 0.229, 0.226, 0.221,
0.210, 0.156, 0.137, 0.131, 0.127, and 0.113,
respectively; all P< .05).
Multivariate stepwise regression analysis in all of the sub-
jects demonstrated that HDL-CE concentrations, PCOS status,

TABLE 2

HII, HDL-related components, and serum MDA levels in patients with different PCOS phenotypes and control women.
PCOS
Controls HA D OA D HA D OA HA D PCO OA D PCO
Variable (n [ 441) PCO (n [ 172) (n [ 91) (n [ 23) (n [ 139)
Age (y) 28.12
4.06 24.86
3.47a 24.65
3.74a 25.91
4.48a 25.64
3.84a
BMI (kg/m2) 21.00
2.73 22.35
3.62a 22.88
4.67a 21.67
3.39 23.63
4.49a
MS, % (n) 5.7 (25) 25.0 (43)a 20.9 (19)a 17.4 (4)a 36.7 (51)a,b,c
IGT, % (n)d 9.2 (24) 28.0 (44)a 17.1 (13) 10.0 (2) 20.9 (27)a
HII 0.53
0.37 0.77
0.52a 0.87
0.59a 0.83
0.43a 0.68
0.53a,c
HII R1, % (n) 8.4 (37) 27.3 (47)a 35.2 (32)a 26.1 (6)a 21.6 (30)a,c
Intrinsic HDL oxidation (rfu/mL) 1.57
0.74 1.48
0.86 1.56
0.87 1.57
0.72 1.31
0.61a
HDL-FC (mmol/L) 0.16
0.07 0.15
0.07a 0.16
0.08 0.16
0.10 0.14
0.07a
HDL-CE (mmol/L) 1.44
0.32 1.36
0.33a 1.39
0.33a 1.50
0.42 1.29
0.35a,e
HDL-TG (mmol/L) 0.27
0.14 0.29
0.16 0.30
0.15 0.26
0.15 0.28
0.15
ApoA-I (g/L) 1.43
0.19 1.41
0.19a 1.38
0.19a 1.40
0.26 1.36
0.19a
MDA (nmol/mL) 3.60
1.21 4.26
1.65a 4.58
1.76a 4.34
1.24a 3.99
1.42a,b,c,e
Note: Values are presented as mean
SD unless otherwise noted. The frequencies of HII R1, MS or IGT were compared by c2 tests. Comparisons of other parameters were corrected for the
differences in age and BMI between the two groups.
a
P< .05 vs. controls.
b
P< .05 vs. the HA þ OA þ PCO subgroup.
c
P< .05 vs. the HA þ OA subgroup.
d
Control group (n ¼ 260), HA þ OA þ PCO subgroup (n ¼ 157), HA þ OA subgroup (n ¼ 76), HA þ PCO subgroup (n ¼ 20), OA þ PCO subgroup (n¼129).
e
P< .05 vs. the HA þ PCO subgroup.
Zhang. Antioxidant properties of HDL in PCOS. Fertil Steril 2015.

1350 VOL. 103 NO. 5 / MAY 2015

 Fertility and Sterility®

TABLE 3

Association of important independent variables with HII or intrinsic HDL oxidation levels by multivariate regression analysis in subjects.
Unstandardized coefficients Standardized coefficients
Independent variable b SE b t P value
a
Model I : HII as dependent
Constant 0.300 0.143 2.096 .037
HDL-CE (mmol/L) 0.496 0.084 0.326 5.934 < .001
PCOS (control ¼ 0, PCOS ¼ 1) 0.300 0.078 0.211 3.865 < .001
E2 (102 pmol/L) 0.027 0.009 0.158 2.911 .004
a
Model II : Intrinsic HDL oxidation levels as dependent (rfu/mL)
Constant 1.439 0.281 5.117 < .001
HDL-CE (mmol/L) 1.322 0.103 0.618 12.833 < .001
FSH (IU/L) 0.047 0.012 0.156 3.838 < .001
ApoA-I (g/L) 0.635 0.193 0.157 3.285 .001
HDL-TG (mmol/L) 0.583 0.201 0.119 2.902 .004
LDL-C (mmol/L) 0.111 0.042 0.109 2.675 .008
F-G score 0.035 0.015 0.092 2.269 .024
a
Each variable in the models includes all of the subjects (PCOS: n ¼ 425, control: n ¼ 441) except for 2-h insulin, 2-h glucose, and IGT status (PCOS: n ¼ 382, control: n ¼ 260).
Zhang. Antioxidant properties of HDL in PCOS. Fertil Steril 2015.

and E2 levels were significant predictors of HII values,
whereas HDL-CE, FSH, apoA-I, HDL-TG, and LDL-C levels
and F-G score were significant predictors of intrinsic HDL
oxidation levels (Table 3).
DISCUSSION
In this study we determined that the decreased antioxidant
activity of HDL and the presence of pro-oxidative HDL were
strongly associated with PCOS. Furthermore, the antioxidant
functions of HDL were more severely impaired in patients
with HA than in those with OA þ PCO.
The absolute risk of the contribution of PCOS status to
cardiovascular disease remains unclear. High TG, low HDL-
C, and increased small LDL-C particles is the most common
metabolic abnormality in women with PCOS (3, 33, 34).
Epidemiologic studies have determined that HDL-C levels
are inversely correlated with a risk of cardiovascular events
(35, 36). However, many patients who experience clinical
events have normal or high levels of HDL-C, suggesting
that HDL levels in these patients may not be protective and
may even promote vascular inflammation and LDL oxidation
(32, 35, 36). Measurement of HDL-C levels may provide infor-
mation about the size of the HDL pool but does not predict its
function (35). One study suggested that HII was a better pre-
dictor for coronary heart disease than HDL-C levels (32).
Oxidative stress is defined as an imbalance caused by
excessive oxidant formation in the presence of limited anti-
oxidant defenses (37). Recently, a systematic review and
meta-analysis demonstrated that the concentrations of
several promoters and by-products of oxidative stress such
as homocysteine and MDA were significantly increased,
and some circulating antioxidant markers, such as PON1 ac-
tivity and glutathione levels, were decreased in patients with
PCOS compared with the control women (4). Our recent
studies indicated that decreased HDL-associated PAF-AH ac-
tivities and sequence variations in antioxidant enzymes PAF-
AH and/or PON1 genes were associated with the increased
PCOS risk (5, 22, 23). In this study we determined that
patients with PCOS had significantly higher HII values, HII
R1 frequency, and serum MDA levels and lower HDL-FC
concentrations than the controls. These results suggested
that the antioxidant/anti-inflammatory function of HDL
was impaired, whereas the increased oxidative stress and
HDL component abnormalities might be associated with
this dysfunction in patients with PCOS. Like LDL, HDL is
susceptible to oxidation (15, 16), and accumulation of
lipid hydroperoxides in HDL may impair the associated
apolipoproteins and enzymes and weaken the athero-
protective abilities of HDL (11, 15, 16, 38). Consistent with
this explanation, Kontush et al. (39) demonstrated that
HDL-mediated protection from lipid oxidation was more
efficient when oxidative stress was low.
Hyperandrogenism is a major characteristic of PCOS.
Hyperandrogenism is often associated with increased visceral
fat and insulin resistance (IR), as well as impaired glucose
and lipid metabolism (40). Our previous study demonstrated
that increased MDA levels were correlated with elevated TT
levels in patients with PCOS (5). In this study we demonstrated
that PCOS phenotypes with HA had higher oxidative stress
and tended to have more severe impairment in the antioxi-
dant function of HDL compared with the non-HA phenotype
(OA þ PCO). We also observed that increased intrinsic
HDL oxidation levels correlated positively with the hirsutism
(F-G) score in a multivariate regression model. Consistent
with our findings, Gonzalez et al. (41, 42) reported that
hyperandrogenemia activated and sensitized leukocytes to
glucose and increased leukocyte reactive oxygen species
generation, NADPH oxidase p47(phox) subunit gene
expression, and plasma thiobarbituric acid-reactive sub-
stances to promote oxidative stress in lean, healthy,
reproductive-age women (41) or in normal-weight PCOS
women (42). Moderate oxidative stress can stimulate prolifer-
ation of theca-interstitial cells and production of T (43).
However, it remains to be determined whether oxidized HDL
also promotes the occurrence of HA.

VOL. 103 NO. 5 / MAY 2015 1351

ORIGINAL ARTICLE: REPRODUCTIVE ENDOCRINOLOGY
Patients with PCOS have increased MS risk, which is asso-
ciated with obesity, increased visceral fat, and dyslipidemia
(3, 44). In this study we demonstrated that patients with OA
þ PCO had higher prevalence of MS but had lower or a
trend to a decrease in HII values, HII R1 frequency, HDL-
FC, HDL-CE, apoA-I, and HDL intrinsic oxidation levels
compared with HA patients. Furthermore, HDL-CE and
apoA-I levels were the main predictors of HDL intrinsic
oxidation levels. The HII values were correlated positively
with intrinsic HDL oxidation levels. The HII values or intrinsic
HDL oxidation levels were negatively correlated with waist-
to-hip ratio, waist circumference, and serum TG levels. These
findings hinted that apoA-I might play an important role not
only in reverse cholesterol transport but also in HDL binding
and transport of oxidized lipids, whereas obesity and/or
dyslipidemia might decrease these capacities of HDL, possibly
resulting from decreased apoA-I levels. Supporting our find-
ings, Roe et al. (33) recently reported that although HDL-C
levels were similar, significant decreases in cholesterol efflux
capacity and apoA-I levels were observed in women with
PCOS compared with controls independent of obesity, and
apoA-I was one of the strongest predictors, although BMI,
fasting insulin levels, HOMA-IR, and TT levels were also asso-
ciated with decreased efflux capacity. Metabolic syndrome
was also characterized not only by low HDL-C and apoA-I
levels but also by defective HDL function (16, 45). Increased
intra-abdominal fat may lower HDL levels by increasing the
catabolic rate of HDL particles without apoA-II (46). Collec-
tively, the effects of obesity and/or dyslipidemia on HDL
functions might be partially related to decreased apoA-I
levels; thus, the weakened abilities of binding and transport-
ing oxidized lipids.
Patients with PCOS also demonstrate increased risk of
IGT. In this study the prevalence of IGT was higher in patients
with HA þ OA þ PCO and OA þ PCO and tended to be
increased in those with HA þ OA compared with the control
women. Insulin resistance is a key pathophysiologic feature of
PCOS (7). It is proposed that women with PCOS have intrinsic
IR that is mechanistically distinct from obesity-associated IR
(47). Insulin resistance can increase HA because insulin in-
creases ovarian androgen production and reduces hepatic
sex hormone–binding globulin production (7, 47); thus it
may impair the antioxidant abilities of HDL through HA
and hyperglycemia by inducing oxidative stress (41, 42).
The increased oxidative stress could impair glucose
metabolism, reduce insulin secretion, induce IR, and
increase IGT risk (4). Our previous findings (5) that PCOS-
intrinsic IR was linked to decreased HDL-associated PAF-
AH activities support this explanation. Our results also
demonstrated that patients with OA þ PCO were more obese
and presented with the highest prevalence of dyslipidemia
and MS (3), as well as a relatively high prevalence of IGT,
suggesting that obesity-associated IR might play a more
important role in IGT in these patients.
We also observed that E2 levels were correlated positively
with HII values or intrinsic HDL oxidation levels in patients
with PCOS. In the multivariate regression analysis that
included all subjects, E2 levels were a significant predictor
of HII values, and FSH was a negative significant predictor
1352
of intrinsic HDL oxidation levels. Several studies demon-
strated that E2 and its 17b fatty acid esters as antioxidants
incorporated HDL particles and significantly increase HDL
antioxidant potential (48, 49). Studies also determined that
menopause impaired HDL antioxidant activity (50), whereas
estrogen (E)/hormone replacement therapy significantly
elevated apoA-I levels and PON1 activity and decreased ho-
mocysteine or oxidized LDL levels in postmenopausal women
(51–53), suggesting that E may enhance the antioxidant
capacities of HDL by elevating apoA-I levels and PON1 activ-
ity. The effect of FSH on the antioxidant function of HDL is
still unclear. However, reduced levels of circulating E result
in sustained high FSH level, increased oxidative stress, and
impaired HDL antioxidant activity in postmenopausal women
(48, 50), suggesting that FSH may affect HDL antioxidant
activity via regulating E levels. In brief, E2 may increase the
ability of HDL binding and oxidized lipid transport by
nongenomic and genomic mechanisms. Conversely, as
mentioned above, increased oxidative stress may result in
excessive accumulation of lipid hydroperoxides in HDL,
which may also impair the associated apolipoproteins and
enzymes and weaken the antioxidant abilities of HDL, thus
increasing HII values. Of note, because E2 has unique
biological effects, the precise mechanism by which E2
affects HDL antioxidant activity still remains to be
investigated further.
There are some limitations to this study. First, we only
measured TT levels, which may be a less sensitive marker
for HA than free T levels. Second, some subjects declined an
oral glucose tolerance test because it is uncomfortable and
time-consuming, so 2-hour insulin, glucose indexes, and
IGT status in these women were not recorded, which might in-
fluence the power of these parameters. Third, in this study the
presence of tubal factor might have biased the control group.
Finally, the HA þ PCO phenotype sample sizes was relatively
small, and a larger number of subjects would be needed to
properly evaluate their characteristics.
In summary, our study provides evidence that the
antioxidant/anti-inflammatory activity of HDL is decreased
in patients with PCOS. In some patients, HDL might even exert
a pro-oxidative/proinflammatory effect. The impairment of
the antioxidative function of HDL in patients with PCOS is
related to HA status, increased oxidative stress, and abnor-
malities in HDL components and thus may contribute to
PCOS pathogenesis. The existence of dysfunctional HDL
and the decrease in HDL-C levels may act synergistically to
accelerate atherosclerosis and increase the risk of cardiovas-
cular disease in these patients.
Acknowledgments: The authors thank the patients with
PCOS and the control women who donated their blood sam-
ples for this study; and You Li, Dehua Wan, Qi Song, and
Feng Zhang for work performed to support this study.
REFERENCES
1. March WA, Moore VM, Willson KJ, Phillips DI, Norman RJ, Davies MJ. The
prevalence of polycystic ovary syndrome in a community sample assessed
under contrasting diagnostic criteria. Hum Reprod 2010;25:544–51.
VOL. 103 NO. 5 / MAY 2015

2. Lim SS, Davies MJ, Norman RJ, Moran LJ. Overweight, obesity and central
obesity in women with polycystic ovary syndrome: a systematic review
and meta-analysis. Hum Reprod Update 2012;18:618–37.
3. Zhang J, Fan P, Liu H, Bai H, Wang Y, Zhang F. Apolipoprotein A-I and B
levels, dyslipidemia and metabolic syndrome in south-west chinese women
with PCOS. Hum Reprod 2012;27:2484–93.
4. Murri M, Luque-Ramirez M, Insenser M, Ojeda-Ojeda M, Escobar-
Morreale HF. Circulating markers of oxidative stress and polycystic ovary syn-
drome (PCOS): a systematic review and meta-analysis. Hum Reprod Update
2013;19:268–88.
5. Fan P, Liu H, Wang Y, Zhang F, Bai H. Apolipoprotein E-containing HDL-
associated platelet-activating factor acetylhydrolase activities and malon-
dialdehyde concentrations in patients with PCOS. Reprod Biomed Online
2012;24:197–205.
6. Repaci A, Gambineri A, Pasquali R. The role of low-grade inflammation in
the polycystic ovary syndrome. Mol Cell Endocrinol 2011;335:30–41.
7. Moran LJ, Misso ML, Wild RA, Norman RJ. Impaired glucose tolerance, type
2 diabetes and metabolic syndrome in polycystic ovary syndrome: a system-
atic review and meta-analysis. Hum Reprod Update 2010;16:347–63.
8. Dokras A. Cardiovascular disease risk in women with PCOS. Steroids 2013;
78:773–6.
9. de Groot PC, Dekkers OM, Romijn JA, Dieben SW, Helmerhorst FM. PCOS,
coronary heart disease, stroke and the influence of obesity: a systematic re-
view and meta-analysis. Hum Reprod Update 2011;17:495–500.
10. Goodarzi MO. Looking for polycystic ovary syndrome genes: rational and
best strategy. Semin Reprod Med 2008;26:5–13.
11. Navab M, Ananthramaiah GM, Reddy ST, Van Lenten BJ, Ansell BJ,
Fonarow GC, et al. The oxidation hypothesis of atherogenesis: the role of
oxidized phospholipids and HDL. J Lipid Res 2004;45:993–1007.
12. Steinberg D, Parthasarathy S, Carew TE, Khoo JC, Witztum JL. Beyond
cholesterol. Modifications of low-density lipoprotein that increase its athero-
genicity. N Engl J Med 1989;320:915–24.
13. Ansell BJ, Watson KE, Fogelman AM, Navab M, Fonarow GC. High-density
lipoprotein function recent advances. J Am Coll Cardiol 2005;46:1792–8.
14. Bergt C, Oram JF, Heinecke JW. Oxidized HDL: the paradox-idation of lipo-
proteins. Arterioscler Thromb Vasc Biol 2003;23:1488–90.
15. Bowry VW, Stanley KK, Stocker R. High density lipoprotein is the major car-
rier of lipid hydroperoxides in human blood plasma from fasting donors.
Proc Natl Acad Sci U S A 1992;89:10316–20.
16. Hansel B, Kontush A, Bonnefont-Rousselot D, Bruckert E, Chapman MJ. Al-
terations in lipoprotein defense against oxidative stress in metabolic syn-
drome. Curr Atheroscler Rep 2006;8:501–9.
17. Morgantini C, Natali A, Boldrini B, Imaizumi S, Navab M, Fogelman AM,
et al. Anti-inflammatory and antioxidant properties of HDLs are impaired
in type 2 diabetes. Diabetes 2011;60:2617–23.
18. Kelesidis T, Currier JS, Huynh D, Meriwether D, Charles-Schoeman C,
Reddy ST, et al. A biochemical fluorometric method for assessing the oxida-
tive properties of HDL. J Lipid Res 2011;52:2341–51.
19. Navab M, Hama SY, Hough GP, Subbanagounder G, Reddy ST,
Fogelman AM. A cell-free assay for detecting hdl that is dysfunctional in pre-
venting the formation of or inactivating oxidized phospholipids. J Lipid Res
2001;42:1308–17.
20. Patel PJ, Khera AV, Jafri K, Wilensky RL, Rader DJ. The anti-oxidative capacity
of high-density lipoprotein is reduced in acute coronary syndrome but not in
stable coronary artery disease. J Am Coll Cardiol 2011;58:2068–75.
21. Fan P, Liu HW, Wan DH, Li Y, Song Q, Bai H. Altered distribution of plasma
platelet-activating factor acetylhydrolase between high-density lipoprotein
and low-density lipoprotein in patients with polycystic ovary syndrome. Fertil
Steril 2009;92:2054–7.
22. Wang Y, Liu H, Fan P, Bai H, Zhang J, Zhang F. Evidence for associa-
tion between paraoxonase 1 gene polymorphisms and polycystic
ovarian syndrome in southwest Chinese women. Eur J Endocrinol
2012;166:877–85.
23. Fan P, Liu HW, Wang XS, Zhang F, Song Q, Li Q, et al. Identification of the
G994T polymorphism in exon 9 of plasma platelet-activating factor acetylhy-
drolase gene as a risk factor for polycystic ovary syndrome. Hum Reprod
2010;25:1288–94.
VOL. 103 NO. 5 / MAY 2015
Fertility and Sterility®
24. The Rotterdam ESHRE/ASRM-Sponsored PCOS Consensus Workshop
Group. Revised 2003 consensus on diagnostic criteria and long-term health
risks related to polycystic ovary syndrome (PCOS). Fertil Steril 2004;81:19–
25.
25. Falsetti L, Gambera A, Tisi G. Efficacy of the combination ethinyl oestradiol
and cyproterone acetate on endocrine, clinical and ultrasonographic profile
in polycystic ovarian syndrome. Hum Reprod 2001;16:36–42.
26. Pochi PE, Shalita AR, Strauss JS, Webster SB, Cunliffe WJ, Katz HI, et al.
Report of the consensus conference on acne classification. Washington,
D.C., March 24 and 25, 1990. J Am Acad Dermatol 1991;24:495–500.
27. Liu HW, Zhang F, Fan P, Bai H, Zhang JX, Wang Y. Effects of apolipoprotein E
genotypes on metabolic profile and oxidative stress in south-west Chinese
women with polycystic ovary syndrome. Eur J Obstet Gynecol Reprod Biol
2013;170:146–51.
28. Guo M, Chen ZJ, Macklon NS, Shi YH, Westerveld HE, Eijkemans MJ, et al.
Cardiovascular and metabolic characteristics of infertile chinese women
with PCOS diagnosed according to the Rotterdam consensus criteria. Re-
prod Biomed Online 2010;21:572–80.
29. Robert Y, Dubrulle F, Gaillandre L, Ardaens Y, Thomas-Desrousseaux P,
Lemaitre L, et al. Ultrasound assessment of ovarian stroma hypertrophy in
hyperandrogenism and ovulation disorders: visual analysis versus computer-
ized quantification. Fertil Steril 1995;64:307–12.
30. Zhang B, Fan P, Shimoji E, Itabe H, Miura S, Uehara Y, et al. Modulating ef-
fects of cholesterol feeding and simvastatin treatment on platelet-activating
factor acetylhydrolase activity and lysophosphatidylcholine concentration.
Atherosclerosis 2006;186:291–301.
31. Rumsey SC, Galeano NF, Arad Y, Deckelbaum RJ. Cryopreservation with su-
crose maintains normal physical and biological properties of human plasma
low density lipoproteins. J Lipid Res 1992;33:1551–61.
32. Ansell BJ, Navab M, Hama S, Kamranpour N, Fonarow G, Hough G, et al. In-
flammatory/antiinflammatory properties of high-density lipoprotein distin-
guish patients from control subjects better than high-density lipoprotein
cholesterol levels and are favorably affected by simvastatin treatment. Circu-
lation 2003;108:2751–6.
33. Roe A, Hillman J, Butts S, Smith M, Rader D, Playford M, et al. Decreased
cholesterol efflux capacity and atherogenic lipid profile in young women
with PCOS. J Clin Endocrinol Metab 2014;99:E841–7.
34. Wild RA, Rizzo M, Clifton S, Carmina E. Lipid levels in polycystic ovary syn-
drome: systematic review and meta-analysis. Fertil Steril 2011;95:1073–9.
35. Navab M, Reddy ST, Van Lenten BJ, Fogelman AM. HDL and cardiovascular
disease: atherogenic and atheroprotective mechanisms. Nat Rev Cardiol
2011;8:222–32.
36. Rothblat GH, Phillips MC. High-density lipoprotein heterogeneity and func-
tion in reverse cholesterol transport. Curr Opin Lipidol 2010;21:229–38.
37. Turrens JF. Mitochondrial formation of reactive oxygen species. J Physiol
2003;552:335–44.
38. Kontush A, Chapman MJ. Functionally defective high-density lipoprotein: a
new therapeutic target at the crossroads of dyslipidemia, inflammation, and
atherosclerosis. Pharmacol Rev 2006;58:342–74.
39. Kontush A, Chantepie S, Chapman MJ. Small, dense HDL particles exert
potent protection of atherogenic LDL against oxidative stress. Arterioscler
Thromb Vasc Biol 2003;23:1881–8.
40. Castelo-Branco C, Steinvarcel F, Osorio A, Ros C, Balasch J. Atherogenic
metabolic profile in PCOS patients: role of obesity and hyperandrogenism.
Gynecol Endocrinol 2010;26:736–42.
41. Gonzalez F, Nair KS, Daniels JK, Basal E, Schimke JM, Blair HE. Hyperandro-
genism sensitizes leukocytes to hyperglycemia to promote oxidative stress in
lean reproductive-age women. J Clin Endocrinol Metab 2012;97:2836–43.
42. Gonzalez F, Sia CL, Shepard MK, Rote NS, Minium J. Hyperglycemia-induced
oxidative stress is independent of excess abdominal adiposity in normal-
weight women with polycystic ovary syndrome. Hum Reprod 2012;27:
3560–8.
43. Kodaman PH, Duleba AJ. Statins in the treatment of polycystic ovary syn-
drome. Semin Reprod Med 2008;26:127–38.
44. Dokras A, Bochner M, Hollinrake E, Markham S, Vanvoorhis B, Jagasia DH.
Screening women with polycystic ovary syndrome for metabolic syndrome.
Obstet Gynecol 2005;106:131–7.
1353
ORIGINAL ARTICLE: REPRODUCTIVE ENDOCRINOLOGY
45. Kontush A, Chapman MJ. Antiatherogenic small, dense HDL–guardian
angel of the arterial wall? Nat Clin Pract Cardiovasc Med 2006;3:144–53.
46. Vajo Z, Terry JG, Brinton EA. Increased intra-abdominal fat may lower HDL
levels by increasing the fractional catabolic rate of Lp A-I in postmenopausal
women. Atherosclerosis 2002;160:495–501.
47. Teede HJ, Hutchison SK, Zoungas S. The management of insulin resistance in
polycystic ovary syndrome. Trends Endocrinol Metab 2007;18:273–9.
48. Bhavnani BR, Stanczyk FZ. Pharmacology of conjugated equine estrogens:
efficacy, safety and mechanism of action. J Steroid Biochem Mol Biol
2014;142:16–29.
49. Hockerstedt A, Jauhiainen M, Tikkanen MJ. Lecithin/cholesterol acyltransfer-
ase induces estradiol esterification in high-density lipoprotein, increasing its
antioxidant potential. J Clin Endocrinol Metab 2004;89:5088–93.
1354
50. Zago V, Sanguinetti S, Brites F, Berg G, Verona J, Basilio F, et al. Impaired
high density lipoprotein antioxidant activity in healthy postmenopausal
women. Atherosclerosis 2004;177:203–10.
51. Topcuoglu A, Uzun H, Aydin S, Kahraman N, Vehid S, Zeybek G, et al. The
effect of hormone replacement therapy on oxidized low density lipoprotein
levels and paraoxonase activity in postmenopausal women. Tohoku J Exp
Med 2005;205:79–86.
52. Sutherland WH, Manning PJ, de Jong SA, Allum AR, Jones SD, Williams SM.
Hormone-replacement therapy increases serum paraoxonase arylesterase
activity in diabetic postmenopausal women. Metabolism 2001;50:319–24.
53. Man RY, Ting LK, Fan S, Lau MM, Siow YL, Chung YH, et al. Effect of post-
menopausal hormone replacement therapy on lipoprotein and homocyste-
ine levels in chinese women. Mol Cell Biochem 2001;225:129–34.
VOL. 103 NO. 5 / MAY 2015