If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
To measure the abdominal subcutaneous fat (SF) and visceral fat (VF) volumes using high-field magnetic resonance imaging (MRI) and to investigate their association with selected anthropometric and biochemical parameters among obese and nonobese apparently healthy participants.
Methods
A cross-sectional study was conducted by recruiting 167 healthy participants. Abdominal scans were acquired at 3T MRI, and the SF and VF were segmented and their volumes were calculated. Selected anthropometric and biochemical measurements were also determined.
Results
A significant difference (P < 0.05) was observed between normal body weight and overweight and obese participants for SF and VF, total abdominal fat volumes, leptin, resistin, adiponectin and waist circumference. Waist circumferences were measured by tape and MRI. Findings revealed that MRI-measured fat volumes were different between males and females and had a significant (P < 0.01) strong positive correlation with body mass index, leptin, resistin and WC and had a negative correlation with adiponectin level. MRI-measured fat volumes were found to correlate moderately with interleukin-6 and weakly with cholesterol, serum triglyceride and low-density lipoprotein. Except for cholesterol, all measured biochemical variables and abdominal fat volumes in the current study were significantly associated with body mass index.
Conclusions
All anthropometric and biochemical parameters showed weak-to-strong associations with the MRI-measured fat volumes. Abdominal fat distribution was different between males and females and their correlations with some lipid profiles were found to be sex dependent. These findings revealed that MRI can be used as an alternative tool for obesity assessment.
Body fat is distributed into 2 main compartments—subcutaneous fat (SF) tissue and visceral fat (VF) tissue. The latter is known to be hormonally active component of total body fat and possesses unique biochemical characteristics that influence several normal and pathological processes in the human body.
Studies on regional adipose tissue distribution and metabolism revealed that the proportion of abdominal adipose (either visceral or subcutaneous) tissue could be a driver of the health risks associated with overweight and obesity.
Abdominal or visceral adiposity is caused by many factors such as increase in the synthesis and secretion of very low-density lipoprotein by the liver; reduced clearance of triglyceride-rich lipoproteins; presence of low-density lipoprotein (LDL) particles and the reduced high-density lipoprotein (HDL) levels.
These adipokines include leptin, adiponectin, resistin, tumor necrosis factor-alpha, interleukin-6 (IL-6), plasminogen activator inhibitor-1, visfatin and chemerin.
Different methods have been used to estimate the body fat distribution. Anthropometric measurements are among these techniques and they are considered as rudimentary and indirect markers for body fat distribution and require the use of many equations that have been validated for different age, sex and ethnic populations.
Validation of skinfold thickness and hand-held impedance measurements for estimation of body fat percentage among Singaporean Chinese, Malay and Indian subjects.
However, the validity of BMI as a measure of obesity has been questioned in different ethnic groups and age populations as it does not measure body fat directly and could not be used as a diagnostic tool for determining fat mass.
On the contrary, BMI can be used as a tool for monitoring and assessing weight status in populations and as a screening tool for identifying potential weight problems among individuals. Other anthropometric measures such as waist circumference (WC), umbilicus circumference and sagittal diameter are considered as easy techniques that could be used to estimate the body fat distribution, but results that come out upon using such techniques are still inconsistent because the body position during measurement may affect the findings.
Several MRI studies have been conducted in an attempt to achieve different goals. Some of these studies were conducted to validate fat tissue segmentation method (manual, semiautomated or fully automated)
The relation between visceral fat measurement and torso level--is one level better than another? The Atherosclerosis Risk in Communities Study, 1990–1992.
Further research attempts were undertaken to assess any possible association between MRI-derived fat volumes and some anthropometric or biochemical parameters,
In these studies, MRI proved to be a valuable and reliable imaging technique to study and assess abdominal fat distribution. In addition, multislice segmentation approach provided a more accurate and consistent readings especially when measuring the VF that varies between slices
and showed a higher level of correlation with total body fat volume as well as anthropometric and biochemical findings when compared to single slice segmentation approach.
Therefore, the main aim of the present study was to measure the abdominal SF and VF volumes (from dual-echo T1-weighted images acquired using a 2-point chemical shift approach) across a range of healthy obese and nonobese individuals and assess any possible association between those volumes, anthropometric and biochemical parameters of those participants to provide a better assessment of obesity using such combined approach of evaluation.
Materials and Methods
Study Population
The present study has been conducted in accordance with The Code of Ethics of the World Medical Association (Declaration of Helsinki). All participants were asked to sign a consent form according to the approval of local research ethics committee and filled out the MRI safety questionnaire before participating in this study. A total of 167 apparently healthy participants (83 males, 84 females; mean age = 27.2 years; age range: 18-51 years) were recruited from the Royal Medical Services personnel (including the security manpower, hospital cleaners and employees in administrative positions) to participate in this cross-sectional study. In addition, other participants who met the inclusion criteria, which require that the participants were disease-free and at least 18 years old, were also recruited. All participants agreed to be enrolled in the present study except for 11 participants, who refused to give a blood sample and complete the anthropometric measurements. The exclusion criteria included being pregnant or lactating (for women) and suffering from eating disorders or having any disease. Sociodemographic and health data were collected by trained research assistant using interview-based questionnaires. The sociodemographic data included age, marital status, household income, education (illiterate, primary and secondary, diploma and BSc and postgraduate degrees), occupation and tobacco usage (current and previous smokers were categorized as smokers and those who never smoked were set as nonsmokers).
Physical activity level was determined using 7-day Physical Activity Recall (PAR).
Participants were asked to respond to a PAR question based on the way they used to behave in their usual days. The number of hours spent in different activity levels were obtained and converted into metabolic equivalents (METs). The total physical activity MET minutes per week was obtained by summing the METs and then performing categorical analysis (inactive, minimally active or health-enhancing physical activity active).
BMI was calculated as the ratio of weight in kilograms to the square of height in meters and was categorized according to the classification system established by the National Institutes of Health: normal body weight, 18.5-24.9; overweight, 25.0-29.9 and obese, >30.0.
Waist circumference (WC) was measured by tape (WC_tape) at the narrowest level between the lowest rib and the iliac crest at the end of normal expiration in standing position.
In addition, WC was measured from the MR images (WC_MRI) at a level just below the lower costal margin using image analysis software (Slice-O-Matic, Tomovision Inc., Montreal, Canada) in a semiautomated approach. All anthropometric measurements were performed by a trained dietitian.
Biochemical Analysis
Venous blood samples were withdrawn after 12-hour overnight fasting by specialized medical laboratory technicians. Serum samples were separated from the whole blood and stored at −80°C until subsequent analyses.
Determination of Adipokines
Serum inflammatory cytokine, IL-6 and obesity-related adipokines, or adipocytokines (adiponectin, resistin and leptin) concentrations were measured by ELX 800 TC models 96-well Elisa Microplate Readers, USA using commercially available enzyme-linked immunosorbent assay (ELISA) kits (RayBio Human IL-6 ELISA Kit, USA; Cat# ELH-IL-6, RayBio Human Acrp30 ELISA Kit USA; Cat# ELH-Adiponectin, RayBio Human Leptin ELISA Kit, USA; Cat# ELH-Leptin, RayBio Human Resistin ELISA Kit, USA and Cat# ELH-Resistin). The reproducibility of the intra-assay and interassay coefficients of variation were <10% and <12%, respectively for all RayBio ELISA kits.
Determination of Lipid Profile
Fasting blood lipid profile, including total cholesterol, LDL, HDL and triglyceride (TG) levels were measured by Jenway 6305 UV/visible spectrophotometer, USA using commercially available enzymatic kits from Teco diagnostics, USA.
MRI Measurement
MR scanning was performed on a 3-T Siemens Trio MR system (Siemens Healthcare, Erlangen, Germany), equipped with 4-channel phase-array body coil. Nonenhanced transverse T1-weighted in-phase and out-of-phase breath-hold spoiled gradient-echo MRI sequence provided by the manufacturer was performed with repetition time of 5 milliseconds, echo time of 1.225 milliseconds (OP imaging) and 2.45 milliseconds (IP imaging), flip angle of 10°, slice thickness of 5 mm, number of slices of 80, matrix size of 256 × 192, field of view of 380 × 285 mm2, number of averages of 1, acceleration factor of 2 and scan time of 18 seconds. The acquired slices covered the region between the level above the diaphragm and the head of the femur in supine position and full expiration. The MRI scanner calculated the “fat” and “water” images as follows:
(1)
(2)
All abdominal MR images were checked by an MRI physicist for any artifact while the participant was still in the scanner. MR fat images of 14 participants were found to be affected by motion (8 cases), motion and phase (3 cases) or phase (3 cases) artifact. Fat-water signal swap artifact was attributed to a failure in phase correction owing to main magnetic field (Bo) inhomogeneity, which was not corrected by scanner׳s shimming at the beginning of scanning. Furthermore, motion artifact was caused by participant׳s breathing during scanning. So, a second scan was performed to re-do the shimming (to correct fat images for the phase error) and to give participants clear breathing instructions to avoid any motion artifact. As a result, all MR fat images of the 14 participants were free of any motion or phase artifacts or both on their second scan.
Image After Processing and Analysis
“Fat” MR images were imported into Slice-O-Matic in their standard formats (Digital Imaging and Communications in Medicine) to segment the SF and VF tissues. Slices covering the region between the top of the diaphragm and the top of the first sacral vertebra (S1) were used to segment the SF and VF tissues. Depending on the height of the participant, a total of 38-46 axial slices per participant were analyzed by a trained technician. Two different methods were used to segment both tissues. While the SF tissue was segmented using snake algorithm, VF tissue was segmented using a threshold-based growing region. Each segmented tissue was saved as a separate “tag” and the total volumes of SF and VF were calculated from all analyzed slices and saved into a separate file for further analysis. The model and method employed to segment the various tissues is fully described and illustrated elsewhere.
Abdominal SF and VF volumes were calculated for all analyzed slices. The Figure shows an example of single axial “fat” MR image and its segmentation into SF and VF tissues in Slice-O-Matic software. The figure displays a 3-dimensional image rendering of both fat tissues (VF and SF) from the top of the diaphragm to the top of the first sacral vertebra.
FIGUREAn example of “fat” MR image in its gray scale (A), segmentation of subcutaneous and visceral fat tissue (B) and 3D rendering of fat tissues from all analyzed slices in different orientations (C and D). 3D, 3 dimensional; MR, magnetic resonance.
All statistical analyses were performed using SPSS 20.0 (SPSS, Chicago, IL). The significance level was set at P < 0.05 (2-tailed). All the data were tested for normality and linearity. The differences in body fat composition (measured with MRI), anthropometric and biochemical parameters according to sex and BMI were evaluated using one-way analysis of variance and least significant difference tests for multiple comparisons. Pearson׳s correlation coefficients using bivariate and partial (adjusted for age, sex, physical activity and smoking) analyses were performed to evaluate all potential associations between body fat composition and BMI with other variables.
Results
The anthropometric, biochemical and lifestyle characteristics of males and females as well as for the whole study samples are shown in Table 1. For participant׳s height, weigh, and WC_tape and WC_MRI, a significant statistical difference (P < 0.05) was detected between males and females with males showing higher values than females. Although the SF tissue volume in males was found to be significantly (P = 0.001) lower (3,779.4 cm3) than in females (5,222.7 cm3), VF tissue volume was significantly (P = 0.001) higher (2,967.1 cm3) in males than in females (1,881.7 cm3). This was also reflected in the ratio of the VF to SF volumes that was found to be significantly (P = 0.000) higher (0.8243) in males than in females (0.3743). No significant difference was found between males and females for adipokines level and lipid profile, except for TG that was significantly (P = 0.001) higher in males (121.4 mg/dL) than in females (84 mg/dL) although it was still within the reference range for both the sexes.
TABLE 1Anthropometric, biochemical and lifestyle differences within the study sample based on sex.
Significant at P < 0.05; BMI, body mass index; HDL, high-density lipoprotein cholesterol; IL-6, interleukin-6; LDL, low-density lipoprotein cholesterol; MRI, magnetic resonance imaging; SEM, standard error of the mean; TG, triglycerides.
P Value
Total
n
Mean
SEM
n
Mean
SEM
n
Mean
SEM
Height (cm)
77
172.4
0.81
81
158.6
0.66
0.001
157
165.3
0.75
Weight (kg)
77
79.1
1.7
81
68.7
1.36
0.001
157
73.8
1.52
BMI (kg/m2)
77
26.7
0.59
81
27.5
0.63
0.255
157
27.1
0.43
Physical activity (MET)
52
23,080.96
2,449.16
63
20,463.81
1,476.72
0.344
115
21,647.22
1,370.41
Waist circumference (cm) by:
Tape
65
93.8
1.4
75
86.8
1.3
0.017
140
90.0
0.99
MRI
83
90.5
1.3
84
83.8
1.4
0.001
167
87.1
0.97
Subcutaneous fat (cm3)
83
3,779.4
220.4
84
5,222.7
273.6
0.001
167
4,505.4
184.1
Visceral fat (cm3)
83
2,967.1
171.3
84
1,881.7
100.2
0.001
167
2,421.1
107.3
Total abdominal fat (cm3)
83
6,746.5
356.4
84
7,104.4
358.4
0.591
167
6,926.5
252.4
Visceral to subcutaneous ratio
83
0.8243
0.0317
84
0.3743
0.1125
0.000
167
0.5953
0.2416
Leptin (ng/mL)
76
7.5
0.45
80
7.3
0.37
0.804
156
7.4
0.29
Resistin (ng/mL)
76
8.6
0.44
80
8.5
0.5
0.956
156
8.5
0.34
IL-6 (pg/mL)
76
9.2
0.5
80
9.5
0.5
0.676
156
9.3
0.35
Adiponectin (pg/mL)
76
11.1
0.5
80
11.3
0.58
0.784
156
11.21
0.38
Cholesterol (mg/mL)
76
163.0
4.3
80
153.3
3.2
0.071
156
158.0
2.7
TG (mg/mL)
76
121.4
8.0
80
84.0
4.7
0.001
156
102.2
4.8
HDL (mg/mL)
76
41.0
1.2
80
42.2
1.0
0.418
156
41.6
0.77
LDL (mg/mL)
76
112.9
3.7
80
106.2
2.9
0.150
156
109.5
2.3
BMI categories, N (%)
Normal
35 (45.5)
35 (43.8)
0.458
70 (44.6)
Overweight
23 (29.9)
20 (25.0)
43 (27.4)
Obese
19 (24.6)
25 (31.2)
44 (28.0)
Marital status, N (%)
Married
31 (40.8)
24 (29.6)
0.053
55 (35.0)
Single
45 (59.2)
53 (65.4)
98 (62.4)
Divorced
–
4 (4.9)
4 (2.5)
Education, N (%)
Illiterate
1 (1.5)
–
0.761
1 (0.6)
Primary and secondary education
14 (20.6)
22 (28.2)
36 (21.4)
Diploma
12 (17.6)
13 (16.7)
25 (14.9)
Bachelor
39 (57.4)
42 (53.8)
101 (61.3)
Master and PhD
2 (2.9)
1 (1.3)
3 (1.8)
Occupation, N (%)
Yes
43 (64.2)
45 (57.7)
0.063
88 (60.7)
No
24 (35.8)
33 (42.3)
57 (39.3)
Current smoking, N (%)
Yes
34 (50.9)
10 (13)
0.004
44 (30.6)
No
33 (49.1)
67 (87)
100 (69.4)
Significant at P < 0.05; BMI, body mass index; HDL, high-density lipoprotein cholesterol; IL-6, interleukin-6; LDL, low-density lipoprotein cholesterol; MRI, magnetic resonance imaging; SEM, standard error of the mean; TG, triglycerides.
Table 2 shows significant (P < 0.05) differences between normal body weight, overweight and obese participants when they were compared for the SF, VF, total abdominal fat volume, leptin, resistin, adiponectin and WC_tape and WC_MRI. Serum levels of IL-6, cholesterol, LDL and HDL were significantly (P < 0.05) greater in obese participants than those reported for overweight and normal body weight participants. TG serum concentration was the only biochemical variable which was significantly higher (P < 0.05) in overweight and obese participants when compared with normal body weight participants.
TABLE 2Anthropometric and biochemical parameters of the study sample based on BMI.
Significant at P < 0.05 with different letters indicating significant differences between groups based on the least significant difference (LSD) test such that, for the same parameter, similar letters indicate no significant difference between the 2 groups of the participants.
P Value
n
Mean
SEM
n
Mean
SEM
n
Mean
SEM
Waist circumference by:
Tape
63
81.8a
0.82
38
90.8b
1.262
38
103.3c
1.3
0.001
MRI
70
77.5a
0.85
42
88.7b
1.2
44
100.2c
1.2
0.001
Subcutaneous fat (cm3)
70
2,793.9a
1,19.2
42
4,526.8b
187.6
44
7,104.4c
315.0
0.001
Visceral fat (cm3)
70
1,535.7a
86.4
42
2,674.8b
203.9
44
3,543.9c
197.6
0.001
Total abdominal fat (cm3)
70
4,329.6a
163.1
42
7,201.6b
257.8
44
10,648.3c
349.7
0.001
Leptin (ng/mL)
68
4.9a
0.18
40
7.9b
0.35
42
11.4c
0.53
0.001
Resistin (ng/mL)
68
6.5a
0.33
40
8.3b
0.37
42
12.5c
0.74
0.001
IL-6 (pg/mL)
68
7.9a
0.46
40
9.2a
0.52
42
11.4b
0.82
0.001
Adiponectin (pg/mL)
68
14.7a
0.4
40
10.4b
0.51
42
6.25c
0.46
0.001
Cholesterol (mg/mL)
68
149.5a
3.5
40
158.8a
5.1
42
171.5b
5.9
0.005
TG (mg/mL)
68
84.2a
5.1
40
109.7b
11.4
42
125.9b
10.2
0.008
HDL (mg/mL)
68
44.5a
1.3
40
40.5a
1.2
42
38.8b
1.3
0.022
LDL (mg/mL)
68
101.0a
2.9
40
109.9a
3.7
42
123.1b
5.5
0.001
BMI, body mass index; HDL, high-density lipoprotein cholesterol; IL-6, interleukin-6; LDL, low-density lipoprotein cholesterol; MRI, magnetic resonance imaging; SEM, standard error of mean; TG, triglycerides.
BMI was categorized according to NIH (21).
Significant at P < 0.05 with different letters indicating significant differences between groups based on the least significant difference (LSD) test such that, for the same parameter, similar letters indicate no significant difference between the 2 groups of the participants.
Findings illustrated in Table 3 revealed that SF, VF and total abdominal fat volumes had significant (P < 0.01) positive correlation with BMI (r = 0.889, 0.640 and 0.900, respectively), leptin (r = 0.863, 0.728 and 0.870, respectively), resistin (r = 0.698, 0.659 and 0.726, respectively), WC_tape (r = 0.778, 0.713 and 0.803, respectively) and WC_MRI (r = 0.950, 0.816 and 0.963, respectively). In contrast, SF, VF and total abdominal fat volume showed a significant (P < 0.01) strong negative correlation with adiponectin level (r = −0.792, −0.727 and −0.817, respectively). Serum IL-6 level was found to have a moderate correlation with the SF (r = 0.533, P < 0.01), VF (r = 0.462, P < 0.01) and total abdominal fat (r = 0.541, P < 0.01). Cholesterol, TG and LDL showed a significant weak positive correlation with SF (r = 0.290, 0.306 and 0.307, respectively, with P < 0.01), VF (r = 0.238, 0.349 and 0.273, respectively, with P < 0.01) and total abdominal fat (r = 0.291, 0.336 and 0.314, respectively, with P < 0.01). An inverse weak correlation was also detected between HDL and VF as well as total abdominal fat (r = −0.335 and −0.274, respectively, with P < 0.05).
TABLE 3Correlation coefficients (r) of subcutaneous, visceral and total fat volumes with anthropometric and biochemical parameters for the study sample.
Parameter
Subcutaneous fat
Visceral fat
Total abdominal fat
Age (years)
r
0.941
0.774
0.943
P
0.001
0.001
0.001
BMI
r
0.889
0.640
0.900
P
0.001
0.001
0.001
Waist circumference (cm) by:
Tape
r
0.778
0.713
0.803
P
0.001
0.001
0.001
MRI
r
0.950
0.816
0.963
P
0.001
0.001
0.001
Leptin (ng/mL)
r
0.863
0.728
0.870
P
0.001
0.001
0.001
Resistin (ng/mL)
r
0.698
0.659
0.726
P
0.001
0.001
0.001
IL-6 (pg/mL)
r
0.533
0.462
0.541
P
0.001
0.001
0.001
Adiponectin (pg/mL)
r
−0.792
−0.727
−0.817
P
0.001
0.001
0.001
Cholesterol (mg/mL)
r
0.290
0.238
0.291
P
0.013
0.044
0.013
TG (mg/mL)
r
0.306
0.349
0.336
P
0.009
0.003
0.004
HDL (mg/mL)
r
−0.229
−0.335
−0.274
P
0.053
0.004
0.020
LDL (mg/mL)
r
0.307
0.273
0.314
P
0.009
0.020
0.007
BMI, body mass index; HDL, high-density lipoprotein cholesterol; IL-6, interleukin-6; LDL, low-density lipoprotein cholesterol; MRI, magnetic resonance imaging; TG, triglycerides.
Table 4 shows a significant positive relationship between the 3 fat volumes (SF, VF and total abdominal fat volumes) and BMI, WC_tape, WC_MRI, leptin and resistin (P < 0.001) in both males and females. Adiponectin showed a significant (P < 0.001) strong negative correlation with SF, VF and total abdominal fat volumes in both males (r = −0.885, −0.819, −0.904, respectively) and females (r = −0.762, −0.732, −0.794, respectively). Additionally, data showed that IL-6 had a significant strong positive correlation with SF (r = 0.627, P = 0.009) and moderate positive correlation with total abdominal fat volumes (r = 0.591, P = 0.016) in males, whereas it had a significant moderate correlation with SF, VF and total abdominal fat volumes (r = 0.477, 0.509 and 0.511, respectively) in females (P < 0.001). Regarding the lipid profile, cholesterol, TG and LDL showed a significant (P < 0.05) positive but weak association with the 3 fat volumes in females. In males, LDL showed a significant moderate association only with SF (r = 0.512, P = 0.042), whereas HDL had an inverse moderate association with the 3 fat volumes (P < 0.05).
TABLE 4Correlation coefficients (r) of subcutaneous, visceral and total fat volumes with anthropometric and biochemical parameters based on sex.
Male
Female
Parameter
Subcutaneous fat
Visceral fat
Total abdominal fat
Subcutaneous fat
Visceral fat
Total abdominal fat
Age (years)
r
0.189
0.561
0.387
0.284
0.447
0.342
P
0.087
0.001
0.001
0.009
0.001
0.001
BMI
r
0.967
0.840
0.968
0.933
0.778
0.939
P
0.001
0.001
0.001
0.001
0.001
0.001
Waist circumference (cm) by:
Tape
r
0.920
0.842
0.936
0.722
0.673
0.747
P
0.001
0.001
0.001
0.001
0.001
0.001
MRI
r
0.972
0.841
0.972
0.943
0.828
0.961
P
0.001
0.001
0.001
0.001
0.001
0.001
Leptin (ng/mL)
r
0.886
0.826
0.907
0.870
0.662
0.858
P
0.001
0.001
0.001
0.001
0.001
0.001
Resistin (ng/mL)
r
0.869
0.720
0.857
0.631
0.657
0.672
P
0.001
0.002
0.001
0.001
0.001
0.001
IL-6 (pg/mL)
r
0.627
0.445
0.591
0.477
0.509
0.511
P
0.009
0.084
0.016
0.001
0.001
0.001
Adiponectin (pg/mL)
r
−0.885
−0.819
−0.904
−0.762
−0.732
−0.794
P
0.001
0.001
0.001
0.001
0.001
0.001
Cholesterol (mg/mL)
r
0.326
0.203
0.297
0.349
0.312
0.357
P
0.218
0.452
0.264
0.010
0.022
0.008
TG (mg/mL)
r
0.464
0.372
0.454
0.321
0.360
0.349
P
0.070
0.156
0.078
0.018
0.007
0.010
HDL (mg/mL)
r
−0.551
−0.555
−0.579
−0.050
−0.186
−0.090
P
0.027
0.026
0.019
0.722
0.179
0.517
LDL (mg/mL)
r
0.512
0.393
0.494
0.293
0.280
0.305
P
0.042
0.133
0.052
0.032
0.040
0.025
Pearson׳s linear correlation coefficients are given at P < 0.05 after adjustment for age, physical activity and smoking.
r, correlation coefficients; BMI, body mass index; HDL, high-density lipoprotein cholesterol; IL-6, interleukin-6; LDL, low-density lipoprotein cholesterol; MRI, magnetic resonance imaging; TG, triglycerides.
Table 5 shows that BMI had a significant strong positive correlation with WC_tape (r = 0.829, P = 0.001) and WC_MRI (r = 0.934, P = 0.001), leptin (r = 0.867, P = 0.001), resistin (r = 0.748, P = 0.001) and IL-6 (r = 0.569, P = 0.001). However, BMI showed a strong inverse correlation with adiponectin (r = −0.837, P = 0.001). In addition, BMI showed a weak positive correlation with TG (r = 0.257, P = 0.03), LDL (r = 0.259, P = 0.028) and weak negative correlation with HDL (r = −0.327, P = 0.005). Similar results were obtained for almost all biochemical variables when males were separated from females. However, serum cholesterol (in males), TG (in males), HDL (in females) and LDL (in females), showed no significant association with BMI.
TABLE 5Correlation coefficients (r) of BMI with subcutaneous, visceral and total fat volumes, and anthropometric and biochemical parameters based on sex.
Parameter
Male
Female
Total
r
P Value
r
P Value
r
P Value
Waist circumference (cm) by:
Tape
0.942
0.001
0.786
0.001
0.829
0.001
MRI
0.972
0.001
0.926
0.001
0.934
0.001
Subcutaneous fat (cm3)
0.967
0.001
0.933
0.001
0.941
0.001
Visceral fat (cm3)
0.840
0.001
0.778
0.001
0.774
0.001
Total abdominal fat volume (cm3)
0.968
0.001
0.939
0.001
0.943
0.001
Leptin (ng/mL)
0.906
0.001
0.864
0.001
0.867
0.001
Resistin (ng/mL)
0.897
0.001
0.698
0.001
0.748
0.001
IL-6 (pg/mL)
0.580
0.018
0.534
0.003
0.569
0.001
Adiponectin (pg/mL)
−0.926
0.001
−0.807
0.001
−0.837
0.001
Cholesterol (mg/mL)
0.283
0.288
0.276
0.043
0.203
0.088
TG (mg/mL)
0.473
0.064
0.277
0.043
0.257
0.030
HDL (mg/mL)
−0.663
0.005
−0.125
0.366
−0.327
0.005
LDL (mg/mL)
0.519
0.039
0.253
0.065
0.259
0.028
Pearson׳s linear correlation coefficients are given at P < 0.05 after adjustment for age, sex, physical activity and smoking.
r, correlation coefficients; BMI, body mass index; HDL, high-density lipoprotein cholesterol; IL-6, interleukin-6; LDL, low-density lipoprotein cholesterol; MRI, magnetic resonance imaging; TG, triglycerides.
Our findings showed that there was a sex difference in abdominal fat distribution with females having a higher proportion of SF and lower VF compared to males although the difference in total abdominal fat in both males and females was insignificant. This could be attributed to the differences in basal fatty acid oxidation, in the regulation of lipolysis and in postprandial fatty acid storage between males and females.
The results of the current study showed that SF, VF and total abdominal fat measured by MRI were strongly and positively correlated with WC_tape and WC_MRI. Some previous studies have reported that WC was correlated better with fat mass and SF than with VF.
attributed that to the amount of SF that was 3.5 times higher than that of VF.
The significant strong correlations that was detected between the 3 fat volumes (SF, VF and total fat) and leptin were in agreement with those found in a study by Lubkowska et al
in which a direct association was found between leptin and all body fat components expressed in absolute values and percentages including VF and SF. In another study by Park et al,
the level of leptin was found to be strongly influenced by SF, whereas the level of adiponectin was significantly affected by VF. In all cases, serum adiponectin level was found to be inversely correlated with SF and VF.
Many studies have produced inconsistent results about the levels of resistin in obesity and especially its etiopathogenetic relationship to insulin resistance.
To our knowledge, this was the first study to show a positive strong correlation between resistin and VF, SF and total abdominal fat.
The positive correlations that were detected between the 3 fat volumes (SF, VF and total fat volume) and IL-6 were in agreement with some other related studies
Comparison of the release of adipokines by adipose tissue, adipose tissue matrix, and adipocytes from visceral and subcutaneous abdominal adipose tissues of obese humans.
in which large portion of circulating IL-6 were found to be produced from both VF and SF depots in obese participants during basal conditions.
The present study showed that cholesterol, TG and LDL were significantly but weakly associated with VF, SF and total abdominal fat. Moreover, a significant weak negative correlation was detected between HDL and the 3 fat volumes. Hoenig et al highlighted the presence of significant (P < 0.05) negative correlation between VF and HDL (r = −0.361) and positive with TG (r = 0.499). However, an inverse association between LDL and VF (r = −0.348) was detected, which indicates a potential discordance between cardiovascular risk and LDL.
conducted a study on 91 apparently healthy obese subjects to assess the associations between serum lipids and specific fat depots at abdominal and hip levels using MRI. In women, an accumulation of VF was associated with a deteriorated lipid profile, even after adjustment for age and body fat percentage; higher TG (P < 0.001), lower levels of HDL cholesterol (P < 0.01) and a diminished HDL cholesterol- to-LDL cholesterol ratio (P < 0.01) compared to those in men.
In men, there was a significant inverse relationship between the abundance of VF and the HDL-to-LDL ratio where significant positive correlations with total LDL cholesterol and TG disappeared after adjustment for age and fat percentage. Within each sex, SF was not significantly related to serum lipids. The above mentioned study concluded that there was a sex difference in the associations between VF accumulation and serum lipids.
In our study, BMI was significantly correlated with SF, VF, total abdominal fat and WC_tape and WC_MRI. Similar results were documented by Janssen et al (2002), in which BMI and WC were found to independently contribute to the prediction of nonabdominal, abdominal SF and VF in white men and women.
Except for cholesterol, all measured biochemical variables in the current study were significantly associated with BMI. These include leptin, resistin, lL-6, adiponectin, TG, HDL and LDL. Arnardottir et al
demonstrated that leptin levels were highly associated with BMI, total abdominal fat and SF than VF volume per se. Many studies reported that adiponectin was negatively correlated with BMI and body fat mass.
showed that serum resistin was significantly higher in obese than in lean volunteers (24.58 ± 12.93 ng/mL; n = 64 versus 12.83 ± 8.30 ng/mL; n = 15; P < 0.01).
Although the 2-point chemical shift MRI approach has been used and validated in many studies, this is the first study, to the best of our knowledge, to use MRI (multislice, semiautomated segmentation), anthropometric and biochemical parameters, collectively, to study the deposition of VF and SF tissues and investigate their effect on obesity in our country. A main limitation of the current study was the relatively small sample size owing to the limited time available to use the MRI scanner at the Royal Medical Services as it was dedicated for clinical imaging only. So, additional studies with a larger number of participants are required to validate these findings.
Conclusions
All anthropometric and biochemical parameters showed weak-to-strong associations with the MRI-measured fat volumes (SF, VF and their total volumes). Abdominal fat distribution was different between males and females and their correlations with some lipid profiles were found to be sex dependent. Based on the above observations and the agreement of these results with those reported in many studies, the 2-point chemical shift MRI approach, used in this study, proved to be a valuable and reliable imaging technique for studying fat deposition and can be used to assess abdominal fat distribution and their relation to anthropometric and biochemical measurements among obese and nonobese healthy participants.
Acknowledgments
The authors would like to thank the Royal Medical Services for their great help and support and for allowing us to use their 3T MRI to scan our volunteers.
References
World Health Organization
Obesity: preventing and managing the global epidemic, Report of a WHO Consultation (WHO Technical Report Series 894).2000;
Validation of skinfold thickness and hand-held impedance measurements for estimation of body fat percentage among Singaporean Chinese, Malay and Indian subjects.
The relation between visceral fat measurement and torso level--is one level better than another? The Atherosclerosis Risk in Communities Study, 1990–1992.
Comparison of the release of adipokines by adipose tissue, adipose tissue matrix, and adipocytes from visceral and subcutaneous abdominal adipose tissues of obese humans.
Present affiliation: Department of Nutrition and Food Technology (RT), Faculty of Agriculture, The University of Jordan, Amman, Jordan; Department of Chemistry (NN), College of Sciences and Health Professions, Cleveland State University, Cleveland, Ohio.
☆The authors have no conflicts of interest to disclose.
☆☆This research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors.
Significant at P < 0.05; BMI, body mass index; HDL, high-density lipoprotein cholesterol; IL-6, interleukin-6; LDL, low-density lipoprotein cholesterol; MRI, magnetic resonance imaging; SEM, standard error of the mean; TG, triglycerides.
30501-8&id=gr1.jpg){kind=link}