Keywords
body composition - trunk fat mass - visceral fat mass - hyperandrogenism - tumor necrosis factor-alpha - interleukin-6
Introduction
Polycystic ovary syndrome (PCOS) is the most common hormonal disorder among women of
reproductive age, affecting approximately 5–10% of this population globally [1]. In addition to chronic anovulation,
hyperandrogenism, and polycystic ovaries, PCOS is often associated with insulin
resistance (IR) [2] and an increased
lifelong risk of cardiovascular risk factors (CVRFs) [3]. Metabolic damage in women with PCOS is
particularly pronounced in those with excess abdominal visceral adipose tissue (VAT)
[4]. This abdominal obesity,
classically related to female hyperandrogenism [5], worsens IR by increasing free fatty acid production, which leads to
defects in postreceptor insulin action [6], and causes an imbalance in proinflammatory and proatherogenic cytokine
production [7]. Indeed, reduced
adiponectin levels and higher circulating levels of many adipokines, cytokines, and
inflammatory markers – including leptin, tumor necrosis factor-alpha (TNF-α),
interleukin-6 (IL-6), plasminogen activator inhibitor-1, soluble intercellular
adhesion molecule-1, and high-sensitivity C-reactive protein – have been
reported in PCOS [8]
[9]
[10], although the role of central fat deposition in these changes has not
been fully established.
Recently, some authors have shown that, in addition to VAT, total body fat and
subcutaneous truncal/abdominal fat also correlate with IR, metabolic parameters, and
hormonal profiles [11]
[12], suggesting that assessing body fat
distribution could help estimate metabolic risk in both sexes, including women with
PCOS. Computed tomography (CT) and magnetic resonance imaging (MRI) are currently
considered the gold standard for distinguishing between intraabdominal and
subcutaneous fat. However, their high costs and the ionizing radiation associated
with CT limit their widespread use [13].
Dual X-ray absorptiometry (DXA) scanning is an accessible technique that is less
costly, highly accurate, and operator-independent. It involves significantly lower
radiation exposure and is a well-validated indirect method for estimating the amount
of visceral fat [14]. DXA studies have
shown increased trunk fat in patients with PCOS compared to controls [4]
[15]; however, whether this central fat depot can be used to screen and
monitor cardiometabolic risk in PCOS remains to be determined [16].
Therefore, we aimed to compare fat distribution between patients with PCOS and body
mass index (BMI)-matched controls, focusing particularly on whether IR, androgen
levels, and inflammatory markers correlate with body composition parameters in PCOS,
using anthropometric adiposity indices and measurements, and DXA.
Subjects and Methods
Patients and study design
Women aged between 18 and 40 years with a BMI<35 kg/m² were recruited from the
outpatient clinics of Pedro Ernesto University Hospital in Rio de Janeiro,
Brazil, and through social media. A BMI≥35 kg/m², pregnancy, chronic diseases,
regular practice of aerobic or anaerobic exercise, history of smoking, alcohol
consumption, or use of anabolic steroids, hormonal contraceptives, or metformin
in the last three months were the exclusion criteria. Further screening based on
medical history, physical examination, and laboratory tests was conducted on the
remaining women. Those who were not eumenorrheic or had androgen levels outside
the normal range were excluded from the control group. Additionally, those who
did not meet the PCOS diagnostic criteria outlined in the 2018 International
Evidence-Based Guidelines, which are based on the Rotterdam Consensus [17], were excluded from the PCOS group.
Hirsutism was defined using the semiquantitative modified Ferriman–Gallwey score
(mFG score)≥6, and biochemical hyperandrogenism was identified by elevated serum
levels of total testosterone (T), delta-4 androstenedione (A4), and
dihydroepiandrostenedione sulfate (DHEAS), and/or an increased free androgen
index (FAI). Chronic anovulation was identified by oligomenorrhea, defined as
fewer than eight cycles per year, or amenorrhea, defined as the absence of
menstrual bleeding for at least 90 days, not due to pregnancy. Using an
ultrasound device with a transducer frequency of≤8 megahertz, polycystic ovarian
morphology was identified by 12 or more follicles measuring 2–9 mm in diameter
and/or an ovarian volume of≥10 ml in at least one ovary. All patients tested
negative for hyperprolactinemia, Cushing’s syndrome, androgen-secreting tumors,
and nonclassical 21-hydroxylase deficiency.
Anthropometry and body composition assessments
Weights and heights were recorded using a digital scale and a stadiometer,
respectively; waist circumference (WC) was measured at the midpoint between the
lower rib margin and the iliac crest, and hip circumference at the level of the
greater trochanters. BMI and waist-hip ratio (WHR) were calculated. Body
composition was assessed using Lunar iDXA equipment (GE Healthcare, Madison, WI,
USA). A single qualified technician conducted DXA scans to measure body fat
percentage, total FM in kilograms (kg), truncal FM (kg), truncal fat percentage,
lean mass [LM (kg)], android and gynoid FM, android and gynoid fat percentages
and estimated VAT (g). The DXA CoreScan mode in the enCORE software, version 17
(GE Healthcare), was used to analyze VAT. The following indices were calculated
based on fat distribution: FM and LM in kg/height in m2 (FM and LM
indices), FM in kg/LM in kg (F/L ratio), truncal FM in kg/legs FM in kg (fat
distribution index, FDI), and the ratio of the percentage of arms and truncal
fat to the percentage of legs fat (upper/lower fat ratio) [18]
[19]
[20].
Laboratory evaluations
Blood samples were collected after an overnight fast, during the early follicular
phase of a spontaneous or induced menstrual cycle, or randomly from patients
with amenorrhea. Serum glucose, glycated hemoglobin (HbA1c), triglyceride (TG),
total cholesterol (TC), and high-density lipoprotein cholesterol (HDL-c) levels
were measured using commercial automated methods. The calculation of low-density
lipoprotein cholesterol (LDL-c) and the homeostatic model assessment for IR
(HOMA-IR) was performed using established formulas. The levels of sex
hormone-binding globulin (SHBG), T, A4, DHEAS, and insulin were measured using
chemiluminescence immunoassay (Immulite 2000, Siemens). TNF-α and IL-6 levels
were measured using an enzyme-linked immunosorbent assay with an ELX800
Microplate reader (BioTek). FAI, lipid accumulation product (LAP), and visceral
adiposity index (VAI) were calculated using the following formulas: FAI=[T
(nmol/l)/SHBG (nmol/l)×100 [21];
LAP=[WC (cm)−58]×TG (mmol/l) [22]; and
VAI={WC (cm)/36.58+[1.89×BMI (kg/m2)]}×[TG (mmol/l)/0.81]×[1.52/HDL-C
(mmol/l)] [23].
Ethical approval
The local ethics committee of Pedro Ernesto University Hospital approved this
cross-sectional study (approval no. 96886918.3.0000.5259). All participants
provided written informed consent after a detailed explanation of the nature,
objectives, and possible risks of the study.
Statistical analysis
Data normality was assessed using the Shapiro–Wilk test. Results were presented
as mean±standard deviation, median (interquartile range), or frequency (n, %) as
appropriate. The unpaired Student’s t-test and Chi-square test were used
to compare continuous and categorical variables between groups, respectively.
Pearson correlation coefficients were calculated to explore the associations
among various measures – metabolic, inflammatory, and hormonal profiles,
anthropometric indices and measurements of central fat accumulation, and DXA
parameters. All analyses were conducted using NCSSM statistical
software (Kaysville, UT, USA), and statistical significance was set at
p<0.05. Sample size calculation was performed using G*Power 3.1.9.4 and
considered the comparison between the two groups {sample size estimation of 54
patients with an actual power of 0.9502120 [t-tests; means: difference
between two dependent means – matched pairs; two-tailed; effect size of 0.50; α
error probability of 0.05; and statistical power (1 – β) of 0.95]} and the
correlation {sample size estimation of 71 patients with an actual power of
0.9521292 [t-tests; correlation: point biserial model; two-tailed; effect
size of 0.40; α error probability of 0.05; and statistical power (1 – β) of
0.95]} employed to investigate our hypothesis further.
Results
In total, 72 women were included (35 PCOS, 37 Controls). [Table 1] presents the clinical and
laboratory characteristics of the groups. Notably, BMI was comparable between the
two groups; however, class I obesity was present in 28.6% of the PCOS patients
compared to 13.5% of the controls, respectively. The PCOS group was younger (p=0.04)
and had significantly higher rates of acanthosis nigricans and familial
hyperandrogenism than the control group. The average mFG score for patients with
PCOS was 10.43±6.60. WC and WHR were significantly higher in the PCOS group.
Table 1 Clinical and laboratory characteristics of the study
groups.
Variables
|
Control (n=37)
|
PCOS (n=35)
|
p-Value
|
Demographic/clinical characteristics
|
Age (years)
|
30.6±6.7
|
28.0±5.3*
|
0.04
|
Race (%)
|
|
|
|
White
|
18 (48.6)
|
11 (31.4)
|
0.32
|
Black
|
2 (5.4)
|
3 (8.6)
|
–
|
Other
|
17 (45.6)
|
21 (60)
|
–
|
Family history of PCOS
|
5 (13.5)
|
13 (37.1)*
|
0.02
|
Modified Ferriman–Gallwey score
|
–
|
10.43±6.60
|
–
|
Weight (kg)
|
65.1±10.9
|
67.8±9.9
|
0.14
|
BMI (kg/m²)
|
25.2±4.2
|
26.6±3.6
|
0.07
|
Normal
|
20 (54.1)
|
12 (34.3)
|
0.17
|
Overweight
|
12 (32.4)
|
13 (37.1)
|
–
|
Class 1 obesity
|
5 (13.5)
|
10 (28.6)
|
–
|
WC (cm)
|
75 [69.5–81.2]
|
80 [74–88]*
|
0.02
|
Waist/hip ratio
|
0.76±0.05
|
0.80±0.06*
|
<0.001
|
VAI
|
2.05 [1.46–3.45]
|
2.95 [2.01–3.56]*
|
0.05
|
LAP
|
12.85 [8.36–21.27]
|
19.30 [13.21–30.90]
|
0.98
|
Acanthosis nigricans
|
3 (8.1)
|
12 (34.3)*
|
<0.001
|
Systolic blood pressure (mmHg)
|
115.8±11.6
|
114.2±9.7
|
0.37
|
Diastolic blood pressure (mmHg)
|
74.5±9.9
|
72.3±9.3
|
0.17
|
Fasting glucose (mg/dl)
|
87.5±8.2
|
87.7±8.5
|
0.90
|
Insulin (mU/l)
|
10.4 [7.4–12.8]
|
13.2 [9.5–16.1]*
|
<0.001
|
HOMA-IR
|
2.24±0.82
|
3.07±1.30*
|
<0.001
|
HbA1c (mmol/mol)
|
30±4
|
33±5*
|
0.03
|
Total cholesterol (mg/dl)
|
172 [149.5–200.5]
|
175 [159–229]
|
0.06
|
HDL-c (mg/dl)
|
52 [[45]–61]
|
47 [[42]–59]
|
0.11
|
LDL-c (mg/dl)
|
101.4 [85.1–123]
|
116 [97–144.2]*
|
0.02
|
Triglycerides (mg/dl)
|
72±25.2
|
84.8±29.2*
|
0.04
|
TNF-α (pg/ml)
|
0.74 [0.60–0.87]
|
0.74 [0.63–0.90]
|
0.11
|
IL-6 (pg/ml)
|
1.28 [0.83–1.75]
|
1.20 [0.96–2.03]
|
0.26
|
T (ng/ml)
|
0.303±0.132
|
0.450±0.141*
|
<0.001
|
SHBG (nmol/l)
|
92.2 [74.9–132]
|
60 [42.6–82]*
|
<0.001
|
FAI
|
1.13 [0.63–1.52]
|
2.64 [1.51–4.15]*
|
<0.001
|
A4 (nmol/l)
|
2.05±0.64
|
2.86±1.08*
|
<0.001
|
DHEAS (μg/dl)
|
138 [114.8–182.5]
|
206 [132–266]*
|
0.02
|
PCOS: Polycystic ovary syndrome; BMI: Body mass index; WC: Waist
circumference; VAI: Visceral adiposity index; LAP: Lipid accumulation
product; HOMA-IR: Homeostasis model assessment of insulin resistance; HbA1c:
Glycated hemoglobin; LDL-c: Low-density lipoprotein cholesterol; HDL-c:
High-density lipoprotein cholesterol; TNF-α: Tumoral necrosis factor-alpha;
IL-6: Interleukin-6; T: Total testosterone; SHBG: Sex hormone-binding
globulin; FAI: Free androgen index; A4: Delta-4 androstenedione; DHEAS:
Dehydroepiandrosterone sulfate; LH: Luteinizing hormone; FSH: Follicle
stimulating hormone; PRL: Prolactin; 17-OHP: 17-Hydroxyprogesterone; TSH:
Thyroid stimulating hormone. *p-Value, unpaired Student’s
t-test or Chi-square test; results expressed as mean±standard
deviation, medians [1st–3rd quartiles] or n (%).
The PCOS group exhibited significantly higher levels of T, A4, FAI, and DHEAS, and
lower levels of SHBG compared to the control group. Fasting glucose levels were
similar; however, insulin, HOMA-IR, and HbA1c levels were significantly elevated in
patients with PCOS. Moreover, differences were observed in TG and LDL-c levels,
while TC, HDL-c, LAP, TNF-α, and IL-6 levels remained consistent between the two
groups. A trend toward increased VAI in patients with PCOS was also observed.
[Table 2] shows the DXA body composition
results for the groups. Patients with PCOS exhibited similar total body fat but had
an excess of truncal FM and a higher percentage of truncal fat compared to the
control group. The LM, LM and FM indices, and gynoid fat levels were similar across
the groups. The F/L ratio was significantly higher in patients with PCOS compared to
controls. The distribution of upper and lower body fat was significantly higher in
the PCOS group, as indicated by the FDI and the upper/lower fat ratio. Women with
PCOS were also characterized by significantly increased android FM, percentage of
android fat, and VAT.
Table 2 Body composition of the study groups assessed using
Dual-energy X-ray absorptiometry.
Variables
|
Control (n=37)
|
PCOS (n=35)
|
p-Value
|
Body fat mass (kg)
|
23.49±8.35
|
26.02±6.65
|
0.07
|
Body fat (%)
|
36.62±7.80
|
39.41±5.80*
|
0.05
|
Trunk fat mass (kg)
|
9.92 [7.23–13.08]
|
13.11 [7.96–16.75]*
|
0.02
|
Trunk fat (%)
|
35.64±9.81
|
40.48±8.40*
|
0.01
|
Lean mass (kg)
|
39.22±5.04
|
39.15±4.53
|
0.49
|
FM index (kg/m²)
|
9.10±3.34
|
10.19±2.68
|
0.06
|
LM index (kg/m2)
|
15.10 [13.88–16.70]
|
15.52 [14.26–16.67]
|
0.19
|
F/L ratio
|
0.601±0.201
|
0.664±0.155*
|
<0.001
|
FDI
|
1.122 [0.934–1.364]
|
1.225 [1.045–1.630]*
|
<0.001
|
Upper/lower fat ratio
|
1.920±0.190
|
2.030±0.269*
|
<0.001
|
Android fat mass (kg)
|
1.41 [0.93–1.95]
|
1.90 [1.26–2.56]*
|
0.02
|
Android fat (%)
|
35.2±11.9
|
41.2±9.7*
|
<0.001
|
Gynoid fat mass (kg)
|
4.55±1.51
|
4.84±1.17
|
0.15
|
Gynoid fat (%)
|
41.9±7.7
|
43.9±5.5
|
0.09
|
Visceral adipose tissue (g)
|
249 [144–518]
|
487 [262–813]*
|
<0.001
|
PCOS: Polycystic ovary syndrome; FM index: Fat mass index; LM index: Lean
mass index; F/L ratio: Fat/lean mass ratio; FDI: Fat Distribution Index.
* p-Value, unpaired Student’s t-test or Chi-square test; results
expressed as mean±standard deviation, medians [1st–3rd quartiles] or n
(%).
[Table 3] presents the associations
between metabolic, hormonal, and inflammatory profiles and anthropometric adiposity
measurements/indices and DXA body composition parameters in the PCOS group. Insulin
and HOMA-IR showed positive correlations with WC, LAP, VAI, total body FM, trunk FM,
and VAT. Although no significant associations were found between androgens and
anthropometry or DXA body composition measurements, SHBG was significantly inversely
associated with BMI, WC, LAP, and total FM, and it also showed a trend toward an
inverse correlation with VAI. A significant positive association was observed
between IL-6 and body FM, trunk FM, F/L ratio, android FM, and VAI. This association
also approached statistical significance with gynoid FM and VAT. Similar
correlations were observed in the control group; however, specific differences were
identified and are detailed in the Supporting Information (Table
1S).
Table 3 Correlations between metabolic, hormonal, and
inflammatory profiles and anthropometric adiposity indices/measurements
and dual X-ray absorptiometry body composition parameters in
PCOS.
|
Fasting glucose
|
Insulin
|
HOMA-IR
|
HbA1c
|
IL-6
|
TNF-α
|
T
|
SHBG
|
FAI
|
A4
|
DHEAS
|
BMI
|
0.08 (p = 0.63)
|
0.37 (p=0.03)
|
0.36 (p=0.05)
|
0.24 (p=0.16)
|
0.17 (p=0.33)
|
–0.23 (p=0.19)
|
–0.14 (p=0.42)
|
–0.47 (p<0.001)
|
0.23 (p=0.18)
|
–0.15 (p=0.39)
|
–0.04 (p=0.79)
|
LAP
|
0.37 (p=0.03)
|
0.59 (p<0.001)
|
0.63 (p<0.001)
|
0.48 (p<0.001)
|
0.32 (p=0.06)
|
–0.08 (p=0.64)
|
–0.13 (p=0.46)
|
–0.39 (p=0.02)
|
0.03 (p=0.87)
|
–0.03 (p=0.86)
|
–0.20 (p=0.25)
|
VAI
|
0.16 (p=0.36)
|
0.40 (p=0.01)
|
0.40 (p=0.02)
|
0.39 (p=0.02)
|
0.37 (p=0.02)
|
0.02 (p=0.91)
|
0.26 (p=0.13)
|
–0.34 (p=0.05)
|
0.18 (p=0.32)
|
0.18 (p=0.29)
|
0.11 (p=0.55)
|
WC
|
0.29 (p=0.10)
|
0.39 (p=0.02)
|
0.39 (p=0.02)
|
0.36 (p=0.03)
|
0.20 (p=0.26)
|
–0.15 (p=0.39)
|
–0.25 (p=0.16)
|
–0.48 (p<0.001)
|
0.18 (p=0.31)
|
–0.10 (p=0.56)
|
–0.00 (p=0.95)
|
Waist/hip ratio
|
0.46 (p<0.001)
|
0.09 (p=0.60)
|
0.14 (p=0.46)
|
0.30 (p=0.07)
|
0.11 (p=0.60)
|
–0.21 (p=0.20)
|
–0.22 (p=0.22)
|
–0.18 (p=0.30)
|
–0.06 (p=0.72)
|
0.13 (p=0.44)
|
–0.11 (p=0.55)
|
Body fat mass (kg)
|
0.36 (p=0.05)
|
0.39 (p=0.02)
|
0.49 (p<0.001)
|
0.31 (p=0.07)
|
0.37 (p=0.03)
|
–0.27 (p=0.12)
|
–0.15 (p=0.40)
|
–0.37 (p=0.03)
|
0.00 (p=0.96)
|
–0.08 (p=0.64)
|
–0.15 (p=0.40)
|
Trunk fat mass (kg)
|
0.36 (p=0.03)
|
0.37 (p=0.03)
|
0.47 (p<0.001)
|
0.36 (p=0.04)
|
0.40 (p=0.02)
|
–0.22 (p=0.20)
|
–0.14 (p=0.41)
|
–0.32 (p=0.06)
|
–0.05 (p=0.79)
|
–0.04 (p=0.81)
|
–0.10 (p=0.57)
|
F/L ratio
|
0.37 (p=0.03)
|
0.28 (p=0.10)
|
0.41 (p=0.02)
|
0.33 (p=0.05)
|
0.38 (p=0.02)
|
–0.34 (p=0.05)
|
–0.04 (p=0.84)
|
–0.25 (p=0.14)
|
–0.12 (p=0.50)
|
0.04 (p=0.80)
|
–0.22 (p=0.19)
|
Upper/lower fat ratio
|
r=0.23; (p=0.19)
|
r=0.19; (p=0.27)
|
r=0.23; (p=0.21)
|
r=0.23; (p=0.19)
|
0.37 (p=0.03)
|
–0.12 (p=0.41)
|
–0.08 (p=0.62)
|
–0.27 (p=0.12)
|
–0.02 (p=0.88)
|
0.01 (p=0.94)
|
0.02 (p=0.91)
|
Android fat mass (kg)
|
0.38 (p=0.02)
|
0.34 (p=0.05)
|
0.44 (p=0.02)
|
0.36 (p=0.03)
|
0.36 (p=0.03)
|
–0.20 (p=0.24)
|
–0.12 (p=0.49)
|
–0.31 (p=0.07)
|
–0.05 (p=0.79)
|
0.02 (p=0.88)
|
–0.06 (p=0.72)
|
Gynoid fat mass (kg)
|
0.27 (p=0.12)
|
0.37 (p=0.03)
|
0.47 (p<0.001)
|
0.28 (p=0.11)
|
0.34 (p=0.05)
|
–0.25 (p=0.15)
|
–0.21 (p=0.22)
|
–0.35 (p=0.04)
|
0.00 (p=0.96)
|
–0.18 (p=0.30)
|
–0.17 (p=0.32)
|
Visceral adipose tissue (g)
|
0.31 (p=0.07)
|
0.41 (p=0.01)
|
0.46 (p<0.001)
|
0.31 (p=0.07)
|
0.33 (p=0.05)
|
–0.16 (p=0.35)
|
–0.10 (p=0.60)
|
–0.24 (p=0.16)
|
0.00 (p=0.97)
|
–0.03 (p=0.84)
|
–0.22 (p=0.19)
|
PCOS: Polycystic ovary syndrome; HOMA-IR: Homeostasis model assessment of
insulin resistance; HbA1c: Glycated hemoglobin; IL–6: Interleukin-6; TNF-α:
Tumor Necrosis Factor-alpha; FAI: Free Androgen Index; A4: Delta-4
androstenedione; T: Total testosterone; SHBG: Sex hormone–binding globulin;
DHEAS: Dehydroepiandrosterone sulfate; BMI: Body mass index; LAP: Lipid
accumulation product; VAI: Visceral adiposity index; WC: Waist
circumference; F/L ratio: fat/lean mass ratio.
In the PCOS group, LAP, VAI, WC, and WHR showed positive correlations with trunk FM
and VAT. Specifically, LAP correlated with r=0.80 and 0.84 (p<0.001); VAI with
r=0.41 and 0.45 (p≤0.02); WC with r=0.84 and 0.75 (p<0.001), and WHR with r=0.43
and 0.45 (p<0.001) (data not shown).
Discussion
In this study, we aimed to evaluate fat distribution in women with PCOS and to
clarify its complex relationship with IR parameters, androgen levels, and
inflammation. The results indicated that women with PCOS exhibit greater truncal and
visceral fat deposition, as well as an unfavorable metabolic profile compared to
BMI-matched controls with similar total body fat. LAP was the most effective
anthropometric parameter for the clinical assessment of women with PCOS.
Furthermore, while biochemical hyperandrogenism did not significantly correlate with
total body fat or its distribution, IR was more closely associated with central fat
deposition than inflammatory cytokines.
PCOS is characterized by central fat deposition, regardless of whether the BMI
indicates normal weight, overweight, or obesity [24]. This excessive abdominal adiposity – which may be associated with
IR, elevated androgen levels, and chronic inflammation – significantly increases
cardiometabolic risk [25]. In line with
our findings, based on DXA measurements, most previous studies have reported similar
LM, an increased F/L ratio, and excessive truncal fat accumulation in women with
PCOS compared to BMI-matched controls. However, in contrast to our results, these
studies also described elevated total FM in these women [18]
[20]. Moreover, an increase in fat deposition in the android region in
normal-weight women with PCOS has been documented [26], aligning with alterations in
anthropometric measures indicative of excess abdominal adiposity, such as higher WC
and WHR. These findings support our results and highlight the usefulness of WC and
WHR measurements for assessing central fat deposition in clinical practice.
Glintborg et al. [18] reported a strong
association of truncal FM, as measured by DXA, with WC in patients with PCOS.
Additionally, they found that truncal FM and WC were equally effective predictors of
IR, serving as useful clinical tools for indirectly detecting this metabolic
dysfunction. In our study, WC, but not WHR, correlated with IR parameters and
exhibited a negative correlation with SHBG, which may indicate IR [27]. The suitability of using simple
anthropometric parameters, such as WC, to predict and monitor visceral obesity and
cardiometabolic risk in women with PCOS was later confirmed [28].
Contextually, LAP and VAI – indicators that combine anthropometric measurements and
blood lipid values – have emerged as important, noninvasive, and cost-effective
markers of IR [29] and VAT dysfunction
[30]. We observed that the PCOS group
tended to have a higher VAI. Moreover, VAI was associated with IL-6, which is
particularly noteworthy as it suggests that VAI may be a more precise indicator of
chronic low-grade inflammation in adipose tissue than LAP. Corroborating our data,
Durmus et al. [31] found that VAI was
higher in overweight and obese women with PCOS than in overweight and/or obese
controls, as well as in normal-weight women with PCOS. Additionally, VAI was
associated with specific metabolic and inflammatory parameters. These findings have
also been reported in lean patients [32]
and may be linked to the severity of anovulation in PCOS [33]. We observed that both VAI and LAP were
positively correlated with truncal FM, VAT, and IR. Furthermore, a significant
inverse relationship between SHBG and LAP was documented, and LAP demonstrated a
stronger correlation with the densitometric measurements of central fat compared to
VAI. The superiority of LAP in estimating central fat depot amounts using the DXA
method has been previously reported [18].
In the present study, both WC and LAP were correlated with DXA-derived central fat
measures and IR markers in women with PCOS. However, the association between IR and
LAP was the strongest among all the anthropometric parameters analyzed in our PCOS
sample, suggesting that LAP should be the primary tool for the clinical assessment
of patients with this condition.
PCOS is a heterogeneous disorder, and the mechanisms underlying IR in this syndrome
are particularly complex because of potential intrinsic IR in muscle and hepatic
tissues, as well as in various fat depots regardless of obesity – although
significantly exacerbated by obesity [34]
[35]. Reportedly, metabolic
parameters and hormone profiles in PCOS are correlated with total body fat, visceral
fat, and subcutaneous truncal/abdominal fat [11]
[12]. The abnormal
morphology and function of adipose cells in PCOS, characterized by changes in
adipocyte size and altered expression and secretion of leptin/adiponectin, are also
potential mechanisms likely linked to elevated androgen levels [36]
[37]. By quantifying FM using DXA, we observed a direct correlation
between total FM, truncal FM, and estimated visceral fat with surrogate markers of
IR in women with PCOS. Puder et al. [38]
reported an inverse correlation between total FM and central/truncal FM, as assessed
by DXA, with insulin sensitivity. This finding is consistent with studies using
established gold standard methods. In a small cohort of overweight women with PCOS,
Barber et al. [39] reported a correlation
between MRI-measured intraabdominal fat and HOMA-IR. Similarly, Penaforte et al.
[40] observed that insulin resistant
women with PCOS exhibited significantly greater amounts of truncal and visceral
abdominal fat than non-insulin resistant women with PCOS – assessed using CT
imaging. Furthermore, a systematic review and meta-analysis of 30 eligible studies
concluded that central FM is elevated in women with PCOS and is directly associated
with increased fasting insulin levels [41].
Some studies have also shown that central fat accumulation is positively associated
with circulating androgen levels, as well as with insulin levels. Employing DXA and
MRI, Dumesic et al. found that androgen levels were significantly higher in patients
with PCOS than in controls, and they observed a correlation between intraabdominal
FM and both fasting insulin and androgen levels [26]. Conversely, free T has been associated with all FM measures assessed
by DXA and identified as an independent predictor of IR [18]. Herein, despite the notably higher
levels of T, A4, FAI, and DHEAS levels, no significant association was observed
between these androgens and the total amount of body fat or its distribution as
measured by DXA, nor with any of the anthropometric indices and measurements.
Therefore, androgens have been shown to be unreliable indicators of body adiposity
in PCOS, consistent with findings from a study that reported excess androgens to be
independent of body fat quantity and distribution, as assessed by DXA in women with
PCOS [38]. Similarly, Tosi et al. [42] studied the body composition of women
with PCOS using DXA and observed that total and central FM were independent
predictors of insulin sensitivity but not of T levels. The varying results suggest
that the impact of androgens on FM in PCOS needs more investigation.
We assessed the status of chronic low-grade inflammation and its association with
body fat distribution, focusing on the measurement of TNF-α and IL-6 levels,
cytokines primarily secreted by inflammatory cells in visceral fat depots [43]. Established data suggest that TNF-α
and IL-6 play roles in the development of obesity-related IR [44]
[45] and may contribute to hyperandrogenism [8] by directly affecting ovarian and
adrenal function [46]. Consequently,
numerous researchers have investigated the associations between these adipocytokines
and PCOS. Similarly to our findings, a meta-analysis by Morreale et al. [47] found no significant differences in the
levels of TNF-α and IL-6 between patients with PCOS and controls. Shroff et al.
[48] evaluated obese women with PCOS
and IR using DEXA and found no statistically significant differences in TNF-α and
IL-6 levels between the PCOS group and a control group matched for total FM.
Conversely, the same study found that total FM was associated with IL-6 and TNF-α
levels, but central FM was not. However, the study did not differentiate between
visceral and subcutaneous fat within the central FM. Our research identified a
direct relationship between IL-6 and various measures of adiposity, including total
FM, truncal FM, F/L ratio, and android FM in PCOS, but only a trend toward an
association with VAT. This further supports the notion that IL-6 could be considered
a potential marker of both total body fat and central fat excess in PCOS, a finding
that aligns with previous research [49]
[50]. However, its role as a
possible indicator of abdominal visceral fat remains uncertain.
The potential limitations of this study are acknowledged. The women with PCOS were
slightly younger on average than those in the control group. However, the age
difference between the groups was marginally significant and not substantial enough
to influence body composition or the development of metabolic and inflammatory
disturbances. The cross-sectional design of this study limits our ability to infer
causality directly. While our assessment of body composition is validated and
accurate, it is not as precise as methods such as MRI or CT imaging. This highlights
the need for additional controlled and prospective studies using DXA, with
age-matched controls and larger sample sizes.
Despite these limitations, our findings indicate that our predominantly overweight
PCOS patients, compared with BMI-matched controls, exhibited elevated levels of
circulating androgens, similar TNF-α and IL-6 concentrations, a greater degree of
IR, and more central fat accumulation, as evidenced by higher WC, WHR, truncal FM,
and VAT, even though their total FM was similar to that of the controls. Among the
anthropometric parameters in the PCOS cohort of the present study, WC and LAP were
useful for assessing IR and both truncal and visceral fat; however, LAP showed a
stronger association with IR markers. This suggests that LAP is a more comprehensive
tool and should be prioritized for routine clinical evaluation of central fat
accumulation and metabolic dysfunction in these women. Finally, our results indicate
that IR, rather than hyperandrogenemia or proinflammatory cytokines, is more closely
associated with central/visceral fat deposition in PCOS.