Abdominal Superficial Subcutaneous Fat

8132019 Abdominal Superficial Subcutaneous Fat

httpslidepdfcomreaderfullabdominal-superficial-subcutaneous-fat 18

Abdominal Super1047297cial Subcutaneous Fat A putative distinct protective fat subdepot in type 2 diabetes

R ACHEL GOLAN RD1

ILAN SHELEF MD MPH2

ASSAF R UDICH MD PHD13

YFTACH GEPNER MPH1

ELAD SHEMESH MD2

YOASH CHASSIDIM PHD4

ILANA HARMAN-BOEHM MD5

YAAKOV HENKIN MD6

DAN SCHWARZFUCHS MD7

SIVAN BEN AVRAHAM RD1

SHULA W ITKOW RD MPH1

IDIT F LIBERTY MD5

OSNAT TANGI-R OSENTAL BA1

BENJAMIN SARUSI MSC2

MEIR J STAMPFER MD DRPH8

IRIS SHAI RD PHD1

OBJECTIVEdUnlike visceral adipose tissue (VAT) the association between subcutaneousadipose tissue (SAT) and obesity-related morbidity is controversial In patients with type 2 dia-betes we assessed whether this variability can be explained by a putative favorable distinctassociation between abdominal super1047297cial SAT (SSAT) (absolute amount or its proportion)and cardiometabolic parameters

RESEARCH DESIGN AND METHODSd We performed abdominal magnetic resonance

imaging (MRI) in 73 patients with diabetes (mean age 58 years 83 were men) and cross-sectionally analyzed fat distribution at S1-L5 L5-L4 and L3-L2 levels Patients completed foodfrequency questionnaires and subgroups had 24-h ambulatory blood pressure monitoring and24-h ambulatory electrocardiography

RESULTSd Women had higher SSAT (37 vs 23 in men P 0001) despite a similarmean waist circumference Fasting plasma glucose (P = 0046)and HbA1c (P = 0006)were bothlower with increased tertile of absolute SSAT In regression models adjusted for age waistcircumference and classes of medical treatments used in this patient population increasedSSAT was signi1047297cantly associated with decreased HbA1c (b = 20317 P = 0013) decreaseddaytime ambulatory blood pressure (b = 20426 P = 0008) and increased HDL cholesterol(b = 0257 P = 0042) In contrast increased percent of deep SAT (DSAT) was associated withincreased HbA1c (b = 0266 P = 0040) and poorer heart rate variability parameters (P = 0030) Although total fat and energy intake were not correlated with fat tissue distribution increasedintake of trans fattendedto be associated with total SAT(r = 0228P =005)andDSAT(r = 020

P = 0093) but not with SSAT

CONCLUSIONSd Abdomina l SAT is composed of two subdepots that asso ciate diff er-ently with cardiometabolic parameters Higher absolute and relative distribution of fat inabdominal SSAT may signify bene1047297cial cardiometabolic effects in patients with type 2diabetes

Diabetes Care 35640ndash647 2012

I

ntra-abdominal visceral adipose tissue(VAT) which is linked to cardiometa-

bolic risk (1ndash3) differs anatomically and functionally from subcutaneous adi-pose tissue (SAT) (4) In addition abdom-inal SAT is further separable into twodistinct subcompartments by the fasciasuper1047297cialis the super1047297cial SAT (SSAT)and the deep SAT (DSAT) (5) Whereasthe abdominal SSAT subdepot is orga-nized into tightly packed lobules thelobules in the DSAT subdepot are largermore irregular and less well organizedand may represent an intermediate tis-sue organization between SSAT and

VAT (4) An increased proportion of VAT (6) isfrequently reported to be associated withtype 2 diabetes and lower heart rate vari-ability an indicator for autonomic neurop-athy and thereforeincreased cardiovascularrisk (7) In contrast with VAT (1ndash3) thereis less consensus regarding the associationbetween abdominal (8) and peripheralSAT with disease risk and both negativeandpositive associations have been reported(9ndash11) Recent studies suggest that abdo-minal DSAT may be associated with diseaseparameters much like VAT particularly

insulin resistance (51213) Yet the as-sociation between peripheral and abdo-minal SAT and cardiometabolic risk isweak if not ldquoparadoxicalrdquo (914) and itis still unknown whether absolute or rel-ative amount of abdominal SSAT is re-sponsible for the putative ldquoprotectiveeffectsrdquo of abdominal SAT reported insome of these studies

RESEARCH DESIGN ANDMETHODS

Study population As part of baseline measurements in the2-year Cardiovascular Diabetes and Eth-anol (CASCADE) randomized controlledtrial a cross-sectional analysis was per-formed in a subgroup of 73 men andwomen aged 41ndash73 years with type 2 di-abetes (de1047297ned as fasting plasma glucose[FPG]126 HbA1c65 physician diag-nosis or evidence of purchase of oral hy-poglycemic medications) who underwentmagnetic resonance imaging (MRI) of theabdomen Persons were excluded if they were smokers pregnant lactating or using

c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c c

From 1The S Daniel Abraham International Center for Health and Nutrition and Department of Epidemi-ology Faculty of HealthSciences Ben-Gurion Universityof the Negev Beer-Sheva Israel the 2Department

of Radiology Soroka University Medical Center Beer-Sheva Israel the 3Department of Clinical Bio-chemistry and The S Daniel Abraham Center for Health and Nutrition Ben-Gurion University of theNegev Beer-Sheva Israel the 4Departments of Physiology and Neuroradiology Soroka University MedicalCenterand ZlotowskiCenterfor Neuroscience Ben-Gurion University of the Negev Beer-Sheva Israel the5Departments of Internal Medicine and Diabetes Soroka University Medical Center Beer-Sheva Israel the6Department of Cardiology Soroka University Medical Center Beer-Sheva Israel the 7Nuclear ResearchCenter Negev Dimona Israel and the 8Channing Laboratory Department of Medicine Brigham and

Womenrsquos Hospital andHarvard MedicalSchool and Departments of Epidemiology and Nutrition HarvardSchool of Public Health Boston Massachusetts

Corresponding author Iris Shai irishbguacilReceived 22 August 2011 and accepted 4 December 2011DOI 102337dc11-1583 Clinical trial reg no NCT00784433 clinicaltrialsgovRG and IS contributed equally to this studycopy 2012 by the American Diabetes Association Readers may use this article as long as the work is properly

citedtheuse iseducationaland notforpro1047297tand the workis notalteredSee httpcreativecommonsorg licensesby-nc-nd30 f or details

640 DIABETES CARE VOLUME 35 MARCH 2012 carediabetesjournalsorg

C a r d i o v a s c u l a r a n d M e t a b o l i c R i s k

O R I G I N A L A R T I C L E



an insulin pump or injecting insulin morethan twice per day had evidence of severediabetes complications (eg proliferativeretinopathy or advanced renal disease)had autonomic neuropathy manifested aspostural hypertension or hypoglycemiaunawareness had a fasting serum triglyc-eride level 400 mgdL or a serum creat-

inine level of $2 mgdL (177 mmolL)had liver dysfunction (greater than twicethe upper limit of normal of alanine ami-notransferase and aspartate aminotransfer-ase levels) had active cancer or receivedchemotherapy in the last 3 years or wereparticipating in another trial The study protocol was approved by the Soroka Uni-versity Medical Center Medical EthicsBoard and Helsinki Committee All partic-ipants provided written informed consent

MRI acquisition and image analysisMRI scans of the abdomen were performedusing a 15 Tesla machine (Intera PhilipsMedical Systems Best the Netherlands)using a body coil Subjects were examinedin thesupineposition witharms positionedparallel along the lateral sides of the bodyMRI scans demonstrating fat in the differ-ent compartments were assessed using aMATLAB-based in-house program (Fig 1)Fat tissues of speci1047297c anatomical land-marks were quanti1047297ed The MRI scan al-lows visualizing the fascia super1047297cialisas a 1047297ne black line To divide SSAT fromDSAT we drew a continuous line over the

fascia super1047297cialis After quanti1047297cation fattissues were divided into color-codedgroups super1047297cial subcutaneous fat =dark blue deep subcutaneous fat = lightblue VAT = green perimuscular fat (fatsurrounding and within the latissimusdorsi and diaphragm) = purple and non-classi1047297ed fat (fat surrounding the vertebraeand fat depots unrelated to any of thegroups listed above) = red Selecting thespeci1047297ed fat mass area was performedusing a semiautomatic method (ie con-nected pixels) or various manual toolssuch as rectanglecirclepolygon or freehand for 1047297ne adjustments and correctionsif needed Quanti1047297cation of the fat massregions included the area of each fat typeand the proportion(percentage) of the totalarea of all fat types To obtain absolutemeasurements in metric units a scalingprocedure was applied before the segmen-tation to determine real pixel dimensionsFinally in accordance with other studies(15) we calculated the fat distribution us-ing the mean of the three slices S1L5L5L4 andL3L2Perimuscular and thenon-classi1047297ed fat tissues totaling a negligible

fraction of total abdominal fat were omittedfrom our analysis All multivariate analyseswere performed twice using the absoluteabdominal fat tissue distribution (SSATDSAT VAT) or the relative (in percent) of each depot from total abdominal adiposetissue (TAAT) (SSAT DSAT VAT)

Clinical parameters Anthropometric measures were evaluated

as clinical estimates of whole-body adi-pose tissue compartments Participantswere weighed without shoes to the nearest01 kg Height was measured using a wall-mounted stadiometer to the nearest mil-limeter for determination of BMI Waistcircumference was measured halfway be-tween the last rib and the iliac crest by thequali1047297ed study nurse with the same typeof measuring tape Mean blood pressurefrom two measures was recorded afterresting with the use of an automated sys-tem (Datascop Acutorr 4 SOMA Technol-ogy Inc Bloom1047297eld CT) Blood sampleswere drawn after at least an 8-h fast andcurrent use of all medication was recordedTo assess blood pressure and heart ratevariability 24-h ambulatory blood pres-sure monitoring (OSCAR 2 oscillometricSunTech Medical Model 222 MorrisvilleNC) and 24-h ambulatory electrocardi-ography (ECG) (LIFECARD-CF DelmarReynolds Medical Ltd Hertfordshire UK)were performed in a substudy group of ourpopulation (n = 31 and n = 37 respec-tively) Heart ratevariability parameters in-cluded time domain variables (16) the

mean duration of the timeinterval betweentwo R waves (RR) graphically presented inthe form of an RR interval tachogram theSD of all normal RR (SDNN) the mean of all the 5-min SDs of NN (normal RR) inter-vals during the 24-h period the root meansquare successive difference calculatesthe square root of the mean of the squareddifferences between successive NN inter-vals over 24 h and the 24-h triangular in-dex the integral of the density distribution

(ie the number of all NN intervals) di-vided by the maximum of the density dis-tribution which is more in1047298uenced by thelower than the higher frequencies Fromthe 24-h blood pressure monitor we calcu-lated the average systolic and diastolicblood pressure during the day (average of records between 6 AM and 10 PM) andnight (average of records between 10 PM

and 6 AM) All medications in current usewere recorded and classi1047297ed

Dietary assessment We evaluated dietary assessment by avalidated food-frequency questionnaire(17) that included 127 food items and 3portion-size pictures for17 items Electronicquestionnaires ensured completeness of thedata by prompting the participant when aquestion was not answered

Statistical analysis We divided our study population acrosstertiles of SSAT (range 5052ndash26836mm2) and used both absolute and pro-portional fat distribution for the analysis

Figure 1dMRI imaging of abdominal fat tissues compartments The subcutaneous fascial planewas delineated using the computer interface semiautomatic method where initially an intensity-based automatic segmentation was generated and presented followed by semimanual 1047297ne tuningThe area of each compartment was quantitated separately Fat tissues of speci 1047297c anatomicallandmarks were quanti 1047297ed and divided into color-coded groups as follows dark blue super 1047297cialsubcutaneous fat light blue deep subcutaneous fat green visceral adipose tissue and red non-

classi 1047297

ed fatd

fat surrounding the vertebrae and fat depots that were unrelated to each of the groupslisted above (A high-quality digital representation of this 1047297gure is available in the online issue)

carediabetesjournalsorg DIABETES CARE VOLUME 35 MARCH 2012 641

Golan and Associates



T a b l e

1 d D i s t r i b u t i o n o f f a t d e p o t s b

i o m a r k e r s a n d c l i n i c a l p a r a m e t e r s a c r o s s t e r t i l e s o f a b s o l u t e a b d o m i n a l S S A T a

m o n g p a t i e n t s w i t h t y p e 2 d i a b e t e s

A b d o m i n a l S S A T

E n t i r e

P

o f t r e n d

L o w t e r t i l e

I n t e r m e d i a t e t e r t i l e

T

o p t e r t i l e

R a n g e 5 0 5 2 ndash 7 2 3 1 m m

2

R a n g e 7 2 8 2 ndash 9 8 9 8 m m

2

R a n g e 9 9 3 0 ndash 2 6 8 3 6 m m

2

R a n g e 5 0 5 2 ndash 2 6 8

3 6 m m

2

S S A T m m

2

6 4 0 9 6

6 2 5

8 5 2 5 6

7 6 9

1 3 4

5 6 6

3 7 9 0

9 4 6 0 6

3 7

7 7

d

A g e y e a r s

5 8 7 6

8 8

5 8 5 6

5 7

5 8 4 6

7 9

5 8 8 6

7 4

N S

M a l e s e x n

( )

2 6 ( 1 0 0 )

1 9 ( 8 6 )

1 6 ( 6 4 )

6 1 ( 8 3 )

0 0 0 1

W e i g h t k g

8 5 2 6

1 1 7

8 6 3 6

1 4 2

8 8 5 6

1 4 1

8 6 6 6

1 3 3

N S

W a i s t c i r c u m f e r e n c e c m

1 0 1 1 6

7 1

1 0 3 7 6

1 0 3

1 0 6 5 6

9 2

1 0 3 8 6

9 2

0 0 1 4

B M I k g m

2

2 8 0 6

3 5

2 8 9 6

3 7

3 1 0 6

3 8

2 9 2 6

3 8

0 0 0 1

S y s t o l i c b l o o d p r e s s u r e m m H g

1 3 8 4 6

1 6 0

1 3 1 1 6

1 4 4

1 3 9 7 6

1 8 3

1 3 6 2 6

1 6 4

N S

D i a s t o l i c b l o o d p r e s s u r e m m H g

8 0 5 2 6

1 0 2

7 8 4 6

9 3

7 8 1 6

1 1 1

7 9 0 6

1 0 1

N S

A b d o m i n a l M R I

T o t a l a b d o m i n a l S A T m m

2

1 2 7 3 3 6

2 9 2 9

1 7 4 6 7 6

3 3 2 5

2 4 0

1 8 6

4 6 5 4

1 8 0 2 4 6

6 0

0 8

0 0 0 1

D e e p a b d o m i n a l S A T m m

2

6 3 2 4 6

2 6 1

8 9 4 1 6

3 1 1 0

1 0 5

6 1 6

3 1 6 9

8 5 6 4 6

3 4

2 9

0 0 0 1

V A T m m

2

1 8 7 8 0 6

5 4 6 0

2 1 4 5 5 6

7 0 5 6

1 7 3

1 4 6

6 8 4 6

1 9 0 8 4 6

6 5

8 3

N S

T A A T m m

2

3 1 5 1 4 6

6 0 0 4

3 8 9 2 2 6

8 6 1 7

4 1 3

3 2 6

7 8 8 0

3 7 1 0 9 6

8 5

6 6

0 0 0 1

F a t t i s s u e s d i s t r i b u t i o n

A b d o m i n a l S S A T p r o p o r t i o n

o f T A A T

2 0 8

2 2 7

3 3 2

2 5 6

d


o f t o t a l a b d o m i n a l S A T

5 2 4

5 0 0

5 6 0

5 2 9

d

A b d o m i n a l S S A T V A T r a t i o

0 3 6 6

0 1

0 4 3 6

0 1

0 8 9 6

0 4

0 5 6 6

0 3

d

B l o o d b i o m a r k e r s

F P G m g d L

1 6 1 1 6

7 5 0

1 4 2 9 6

2 9 7

1 2 7 8 6

2 3 0

1 4 3 7 6

4 8 9

0 0 4 4

H b A 1 c

7 9 6

1 4

7 5 6

1 0

7 0 6

0 4

7 5 6

1 1

9

0 1 0

S e r u m T G m g d L

1 3 6 5 6

6 3 0

1 7 9 5 6

8 4 8

1 6 6 0 6

8 3 6

1 6 0 8 6

7 8 8

N S

S e r u m H D L c h o l e s t e r o l

m g d L

4 5 0 6

6 6

4 1 6 6

1 1 6

4 8 9 6

1 0 2

4 5 1 6

1 0 0

N S

S e r u m L D L c h o l e s t e r o l m g d L

9 2 7 6

3 1 4

9 7 4 6

3 0 5

1 0 1 6 6

2 7 8

9 6 8 6

2 9 9

N S

P l a s m a a l k a l i n e p h o s p h a t a s e u n i t s L

7 0 4 6

2 5 1

8 1 0 6

3 4 8

6 3 1 6

1 5 4

7 2 3 6

2 7 8

N S

P l a s m a a l a n i n e a m i n o t r a n s f e r a s e u n i t s L

2 5 2 6

1 0 0

2 8 9 6

1 1 1

2 6 2 6

1 1 8

2 6 8 6

1 0 9

N S

M e d i c a t i o n s i n c u r r e n t u s e

I n s u l i n t h e r a p y n

( )

3 ( 1 1 )

5 ( 2 3 )

2 ( 8 )

1 0 ( 1 4 )

N S

O r a l g l y c e m i c - c o n t r o l m e d i c a t i o n s n

( )

1 7 ( 6 5 )

1 8 ( 8 2 )

1 6 ( 6 4 )

5 1 ( 7 0 )

N S

L i p i d - l o w e r i n g t h e r a p y n

( )

1 3 ( 5 0 )

1 5 ( 6 8 )

1 6 ( 6 4 )

4 4 ( 6 0 )

N S

A n t i h y p e r t e n s i v e t h e r a p y n

( )

1 3 ( 5 0 )

1 4 ( 6 3 )

1 4 ( 5 6 )

4 1 ( 5 6 )

N S

O r a l a n t i p l a t e l e t d r u g s n

( )

1 2 ( 4 6 )

1 1 ( 5 0 )

1 2 ( 4 8 )

3 5 ( 4 8 )

N S

D a t a a r e m e a n s 6

S D u n l e s s o t h e r w i s e i n d i c a t e d n = 2 6 2 2 a n d 2 5 f o r l o w i n t e r m e d i a t e a n d

t o p t e r t i l e s r e s p e c t i v e l y T G t r i g l y c e r i d e s

P o

f t r e n d

0 0 5 ( A N O V A ) T o c o n v e r t v a l u e s f o r

c h o l e s t e r o l t o m i l l i m o l e s p e r

l i t e r m u l t i p l y b y 0 0 2 5 8 6 T o c o n v e r t v a l u e s f o r t r i g l y c e r i d e s t o m i l l i m o l e s p e r l i t e r m u l t i p

l y b y 0 0 1 1 2 9 T o c o n v e r t v a l u e s f o r g l u c o s e t o m i l l i m o l e s p e r l i t e r m u l t i p l y b y 0 0 5 5 5 1


Super 1047297cial subcutaneous fat depot



to account for both subdepot adiposity and interdepot distribution respectivelyFat distribution was calculated by divid-ing each fat depot by TAAT creating threenewvariablesSSATDSAT andVAT

ANOVA linear test was used to evaluatethe characteristics of the study popula-tion across SSAT tertiles We performed

multivariate linear regression models adjusted for age and waist circumference toevaluate associations among SSAT anthro-pometric measures diet blood biomark-ers 24-h ambulatory blood pressuremonitoring and24-h ambulatory ECG re-cordings Models including cardiovascu-lar outcomes (24-h ambulatory bloodpressure monitoring and 24-h ambulatory ECG recordings) were further adjustedone at a time for FPG or HbA1c We fur-ther performed similar models adjustedone at a time to the various classes of medical treatments used in this patientpopulation (insulin oral hypoglycemicmedications antihypertensive medica-tions and lipid-lowering medications)and performed the same models strati1047297edby sex The Statistical Package for the SocialSciences (version19 SPSS Inc Chicago IL)was used for all statistical analysesP 005denoted statistical signi1047297cance P value wasnot adjusted for multiple testing Values re-ported are means 6 SDs unless otherwisestated Multiple linear regression results arereported with the parameter estimate andP value for each variable

RESULTS

Fat distribution blood biomarkersand dietThe baseline characteristics of the partici-pants in the entire study group and acros-stertiles of SSAT areshown in Table 1 MeanHbA1c was 75 6 11 and 10 (14) pa-tients were taking one daily dose of insulinThe oral hypoglycemic medications used by the patients included sulfonylureas (11patients15) dipeptidyl peptidase-4 in-hibitor (9 patients 12) and metformin(31 patients 42) None of the patientswere taking thiazolidinediones The meanfat tissuesdistribution was as followsSSAT26 DSAT 23 and VAT 51 SSATpositively correlated with DSAT (r =0389 P 0001) and TAAT (r = 0461P 0001) SSAT DSAT and TAAT cor-related positively and signi1047297cantly withwaist circumference (SSAT r = 0313 P =0009 DSAT r = 0276 P = 0023 TAATr = 0449 P 0001) and BMI (SSAT r =0490 P 0001 DSAT r = 0327 P =0005 TAAT r = 0508 P 0001) VAT

correlated positively and signi1047297cantly withwaist circumference and weight (r = 0330P = 0006r = 0346P = 0003respectively)Medical treatment includinginsulin therapy and antihypertensive lipid-lowering anti-platelet and oral hypoglycemicmedicationswas similarly distributed across SSAT tertiles(Table 1)

FPG (P of trend = 0046 P = 0073between extreme tertiles) and HbA1c (P of trend = 0006 P = 0011 between extremetertiles) were both lower with increasedtertile of SSAT SSAT was inversely corre-lated with HbA1C (r =20262 P = 0027)whereas VAT (r = 0240 P = 0042) andTAAT positively correlated with fastingtriglycerides (r = 0278 P = 0018)

In a regression model (Fig 2 A) adjusted for age and waist circumferenceincreased absolute SSAT was associatedwith decreased HbA1c (b = 20289 P =0017) and tended to be associated withdecreased fasting glucose (b = 202080084) Further adjustment of the modelto medical treatment classes (one at atime insulin therapy and antihyperten-sive lipid-lowering antiplatelet andoral hypoglycemic medications) did notsigni1047297cantly attenuate these associations(data not shown) Total energy intake to-tal dietary fat saturated fat and unsatu-rated fat content were not associated withfat tissue mass (data not shown) How-ever increased total dietary intake of transfatty acids tended to be positively corre-

lated with total SAT (r = 0228 P = 0054)and DSAT (r = 0200 P = 0093) Collec-tively the absolute area (mass) of abdom-inal SSAT seemed to be associated withmore favorable glycemic and cardiovas-cular parameters unlike DSAT or VAT

Proportional fat tissue distribution We next assessed whether the relative adi-pose tissue distribution ie the percentageof fat in a certain abdominal (sub)-depotwas associated with clinical parametersThe percentage of SSAT from TAAT(SSAT) was associated with lowerHbA1c in regression models adjusted forage and waist circumference (b =20304P = 0017) consistent with the 1047297ndingusing the absolute SSAT area mentionedabove When we added medical treat-ment to the model (insulin therapy andantihypertensive lipid-lowering antipla-telet and oral hypoglycemic medications)the association remained signi1047297cant Con-versely increased DSAT was associatedwith increased levels of HbA1c (b =0266 P = 0039) and fasting glucose(b = 0246 P = 0054) Increased SSAT

was associated with increased HDL cho-lesterol (b = 0251 P = 0047) (Fig 3 A)

Of the 12 women recruited 11 werepostmenopausal none of whom weretaking hormone replacement therapy

Although men and women had similarwaist circumferences (104 vs 103 cmP =0902) TAAT (37095 vs 37108 mm2P = 0975) and medical treatmentwomen had higher SSAT than didmen and twice the ratio of SSAT to VAT(Table 2) They had lower fasting glucose(P 0001)and lower HbA1c (P = 0002)and tended to have higher HDL choles-terol (P = 0076) Men had a higher VATto SAT ratio whereas DSAT was similarbetween men and women To verify thatthe association between SSAT and favor-able cardiometabolic parameters did notsimply re1047298ect intrinsic differences betweengenders we strati1047297ed theregression models

Figure 2d Association of metabolic parame-ters and cardiovascular parameters with ab-solute abdominal fat tissues in patients withtype 2 diabetes strati 1047297ed by sex A Entiregroup B Men only Multivariate model adjusted for age and waist circumference SSAT (black bar) DSAT (white bar) Numbers rep-resent b standardized coef 1047297cient the amountand direction by which absolute abdominal fat

tissues change (mm 2

) for eachunit changein themetabolic parameters while accounting for theother variables in the model TG triglycerides(mgday) P 005 daggerP 01





by sex (Figs 2B and 3B) In a subgroup of the men only similar opposite associationsamong SSAT DSAT and cardiometa-bolic parameters were observed (Fig 3B)In addition in the men-only subgroup DSAT was signi1047297cantly associated withhigher glucose (b = 0292 P = 0040) andHbA1c (b = 0298 P = 0038)

Fat distribution 24-h ambulatoryblood pressure monitoring and24-h ambulatory ECG

We further found associations among theabsolute abdominal subdepot fat (mass)proportional abdominal subdepot fat areablood pressure and heart rate variability parameters Both absolute (r = 20381P = 0050) and proportional DSAT (r =20428 P = 0026) negatively correlatedwith SDNN In regression models ad-

justed for age and waist circumferencehigher absolute DSAT (b = 20425 P =0034)and higherproportional DSAT (b =20431 P = 0029) were associated with

lower SDNN Proportional DSAT was sig-ni1047297cantly associated with decreased 24-htriangular index (b =20417 P = 0030)suggesting decreased heart rate variability with increased fat distribution to theDSAT depot However further adjust-ment to oral hypoglycemic medicationsbut not to insulin therapy antihyperten-sive lipid-lowering or antiplatelet drugsattenuated the association between DSATand markers of heart rate variability (24-htriangular index b = 20249 P = 0228SDNN b = 20294 P = 0186) Becausehyperglycemia is a key risk factor for au-tonomic neuropathy we further adjustedthe models one at a time for HbA1c orFPG The association between both abso-lute and proportional DSAT and heart ratevariability remained negative but the sig-ni1047297cance was attenuated When adjustedfor HbA1c higher absolute (b = 20355

P = 0093) and proportional (b = 20401P = 0073) DSAT were associated withlower SDNN Proportional DSAT wasneg-atively associated with decreased 24-htriangular index (b = 20406 P = 0063)

When adjusted for glucose higher absolute(b = 20278 P = 0144) and proportional(b = 20322 P = 0122) DSAT were notassociated with lower SDNN In contrastopposite trends were observed with absoluteand proportional abdominal SSAT Higherabsolute (b =20329 P = 0043) and pro-portional SSAT (b = 20425 P = 0008)were associated with lower mean daytimediastolic blood pressure When medicaltreatment was added to the model (insulintherapy antihypertensive lipid-loweringantiplatelet or oral hypoglycemic medica-tions) the association remained signi1047297-cant Increased proportional SSAT tendedto positively correlate with increased

Table 2d Abdominal fat tissue distribution biomarkers and clinical parameters

in men and women with type 2 diabetes

Men Women

Age years 584 6 75 597 6 72

Weight kg 877 6 131 815 6 134

BMI kgm2 289 6 37 312 6 38

Systolic blood pressure mmHg 1349 6 145 1440 6 250Diastolic blood pressure mmHg 787 6 104 807 6 87

Waist circumfere nce cm 1040 6 92 1035 6 95

Abdominal MRI

Total SAT mm2 17080 6 5312 22825 6 7230

Abdominal SSAT mm2 8587 6 2673 13898 6 5357dagger

Abdominal DSAT mm2 8493 6 3403 8927 6 3691

VAT mm2 20014 6 6629 14355 6 3838

TAAT mm2 37095 6 8848 37180 6 7290

Fat distributions

Abdominal SSAT total adipose tissue ratio 022 6 005 036 6 009dagger

Abdominal SSAT total SAT ratio 051 6 01 060 6 01

Abdominal SSAT VAT ratio 047 6 02 105 6 05dagger

VAT total abdominal SAT ratio 128 6 05 069 6 03dagger

Fasting blood biomarkers

Plasma glucose mgdL 1446 6 403 1182 6 118

HbA1c 76 6 11 69 6 04

Serum TG mgdL 1622 6 832 1540 6 540

Serum HDL cholesterol mgdL 441 6 101 500 6 82

Serum LDL cholesterol mgdL 945 6 297 1088 6 293

TGHDL 38 6 23 34 6 19

Plasma alkaline phosphatase unitsL 738 6 287 604 6 148

Plasma alanine aminotransferase unitsL 2736 113 235 6 77

Medications in current use

Lipid-lowering therapy n () 36 (59) 8 (66)

Antihypertensive therapy n () 32 (52) 9 (75)

Oral glycemic-control medications n () 42 (69) 9 (75)

Oral antiplatelet drugs n () 28 (46) 7 (58)Data are means 6 SD unless otherwise indicated n = 61 and 12 for men and women respectively TGtriglycerides P 005 daggerP 0001 To convert values for cholesterol to millimoles per liter multiply by 002586To convertvaluesfor triglycerides to millimolesper liter multiplyby 001129 To convertvaluesforglucose to millimoles per liter multiply by 005551

Figure 3d Association of metabolic and car-diovascular parameters with abdominal fattissues distribution in patients with type 2 dia-betes strati 1047297ed by sex A Entire group B Menonly Multivariate model adjusted for age andwaist circumference SSAT (black bar)DSAT(white bar) Numbers represent b standardizedcoef 1047297cient the amount and direction by which proportional abdominal fat tissues change () for each unit change in the metabolic param-

eters while accounting for the other variables inthemodel TGtriglycerides (mgday) P005





square root of the mean of the squareddifferences between successive NN inter-vals over 24 h (r = 0367 P = 0060)

When HbA1c was added to the modelthe association between absolute SSATand mean daytime diastolic blood pressurewas attenuated (b = 20284 P = 0072)although the association between propor-

tional SSAT and mean daytime diastolicblood pressure remained signi1047297cant (b =20379 P = 0008) When adjusted forglucose the association between both ab-solute (b = 20329 P = 0043) and pro-portional (b = 20444 P = 0006) SSATand mean daytime diastolic blood pres-sure remained signi1047297cant

When the models were strati1047297ed by sex we found that the men-only sub-group also exhibited similar associationsamong abdominal subdepot fat (mass andproportion) blood pressure and heartrate variability parameters Higher abso-lute DSAT (b = 20452 P = 0014) andproportional DSAT (b = 20535 P =0009) were associated with lower SDNNHigher absolute DSAT (b = 20486 P =0007) and proportional DSAT (b =20557 P = 0005) were associatedwith decreased 24-h triangular indexFurther adjustment to oral hypoglyce-mic medications but not to insulin ther-apy antihypertensive lipid-lowering orantiplatelet drugs attenuated the associ-ation between proportional DSAT andmarkers of heart rate variability (24-h tri-

angular index b = 20439 P = 0075SDNN b = 20443 P = 0074) and be-tween absolute markers of heart rate vari-ability (24-h triangular index b =20408P = 0073 SDNN b =20408 P = 0072)

CONCLUSIONSdIn this study of pa-tients with type 2 diabetes we observed adistinct association between both theabsolute (representing ldquosub-depot adi-posity rdquo) and the relative amounts (repre-senting interdepot distribution) of abdominal SSAT and markers of more fa-vorable glycemic control and cardiovascu-lar function as determined by higher heartrate variability and lower blood pressure

After adjusting for age and waist circum-ference higher relative distribution of ab-dominal fat in SSAT was correlated withimproved glycemic control (HbA1c andfasting glucose) and better indicators of cardiovascular health (lower blood pres-sure and higher heart rate variability)Conversely DSAT correlated with higherheart rate and lower heart rate variabil-ity both indicators of autonomic neurop-athy (18) and therefore indicators for

increased cardiovascular risk Of notecontrolling for markers of glycemic con-trol attenuated the negative association be-tween DSAT and cardiovascular endpoint but not the favorable associationwith SSAT In terms of diet increased in-take of trans fat tended to be associatedwith total abdominal SAT and DSAT but

not with SSAT Because most studies sup-port the more ldquoadverse metabolic rolerdquo of intraabdominalVAT whether SAT is sim-ply ldquoless pathogenicrdquo than VAT or exertsdirect or indirect ldquoprotective effectsrdquo oncardiovascular and metabolic morbidity is still controversial (69) Furthermoreeven more uncertainty exists as to thefunctional differences and risk associatedwith the DSAT or SSAT subdepots in per-sons with type 2 diabetes (19) Our resultssuggest a favorable distinct association be-tween abdominal SSAT subdepot and car-diometabolic health in type 2 diabetes

Our study has several limitationsThis is a cross-sectional design that may suggest associations but not clear causendash

effect protective relationships betweenabdominal SSAT and cardiometabolic pa-rameters Because we did not measure pe-ripheral (lower body) SAT our resultsrefer to abdominal SSAT only Our samplesize limits the statistical power althoughwe still could identify signi1047297cant differ-ences between parameters A signi1047297cantproportion of our patients received treat-ment with medications that may modify

the levels of risk factors and directly affectglycemic control parameters Although wehave made an attempt to adjust our resultsto the use of these drugs doses were notassessed Finally we had a lower propor-tion of women but nevertheless we couldidentify signi1047297cant differences between thesexes The strengths of our study includethe speci1047297c group with type 2 diabetes thehigh MRI quality imaging the compre-hensive 24-h ambulatory blood pressureand 24-h ambulatory ECG measurementsas cardiometabolic measurements themedication follow-up and the dietary assessment

We found that abdominal SSAT cor-related with improved glycemic controland indicators of cardiovascular risk TheSSAT depot may be less lipolytic than

VAT or even DSAT and so improved in-sulin sensitivity of SSAT may favor accu-mulation of excess energy in this depot Inthis regard higher deposition of excesscalories in the SSAT is a consequence notthe cause of improved metabolic func-tion Conversely it is plausible that theabdominal SSAT fat mass may be a unique

abdominal fat subdepot that has protec-tive effects on glycemic control and car-diovascular function This is reminiscentof a 1047297nding by some but not all studiesthat suggest peripheral SAT might be lessldquopathogenicrdquo than VAT Currently twohypotheses have been put forward to ex-plain the difference between peripheral

SAT and VAT The ldquoportal theory rdquo (620)implicates a direct mechanism whereby

VAT is more pathogenic because its ve-nous blood drainage is directly via the por-tal vein to the liver The ldquoectopic fathypothesisrdquo (6) suggests an indirect mech-anism whereby increased energy storagein peripheral SAT exerts a protective effectby decreasing fat deposition in the livermuscle and heart Because abdominalDSAT exhibits an intermediate phenotypebetween VAT and abdominal SSAT in var-ious functions tested (lipolysis adipocyto-kine pro1047297le) (4) it is possible that thesetheories can underlie the unique positiveassociation between abdominal SSAT andcardiovascular and metabolic health Al-though it may be possible that peripheralSAT differs signi1047297cantly from abdominalSAT and some studies have indicated thepotential pathogenic role of increasedabdominal (total) SAT as opposed to pe-ripheral SAT (9ndash11) further studies arerequired to fully understand the distinctrole of the SSAT subdepot It is temptingto speculate that discrepancies in the liter-ature among studies assessing associations

between abdominal SAT and morbidity were con1047297ned by differences in the SSATor SSATDSAT distribution

Nutritional habits may also affect fatdistribution In our study increased trans

fatty acids consumption tended to be asso-ciated with increased total abdominal SATand DSAT but not SSAT Limited evidencehas suggested that increased dietary intakeof trans fatty acids may increase fat accu-mulation and abdominal circumference(2122) Although these studies emphasizethe increase in VAT our 1047297ndings suggestthat DSAT also may be increased by exces-sive intake of trans fatty acids

Women in our study had a higherproportion of abdominal SSAT whereasmen had a higher intra-abdominal fat tosubcutaneous fat ratio The proportion of DSAT was similar in men and womenSimilar 1047297ndings have been shown by other cohorts (122324) Fat is distrib-uted in a sexual dimorphic manner be-yond the differences between upper body (more in men) versus lower body loca-tions (more in women) For a givenamount of intra-abdominal fat women





who are relatively protected from cardio-metabolic morbidity at least before men-opause (6) possess up to twice as muchsubcutaneous fat as men (2526) Of notealthough sexual dimorphism in fat distri-bution usually tends to diminish withpostmenopausal state we observed suchchanges in our cohort despite the fact that

most (1112) women were postmeno-pausal This 1047297nding may complement aprior report that postmenopausal womenwho possess a higher proportion of adi-pose tissue located in the total midthighdepot had a more favorable metabolic pro-1047297le (25) Although studies have reported anassociation between hormone replace-ment therapy and cardiovascular health(2728) none of the women in our study were on such therapy Future studies willunravel whether increased abdominalSSAT and its seemingly favorable meta-bolic and cardiovascular correlates dimin-ish with years into menopause or haveother determinants possibly unique topersons with diabetes

Finally it has been proposed that anti-diabetic medications particularly of thethiazolidinedione family may exert theirtherapeutic effect at least partially by in-ducing redistribution of fat from ldquopatho-genicrdquo to ldquoless-pathogenicrdquo depots (egfrom VAT to SAT) (2930) Howevernone of the patients in our study were re-ceiving thiazolidinedione therapy Otherhypoglycemic medications were distrib-

uted similarly among tertiles of SSAT andcontrolled for in the regression models

In summary the current study addsto the understanding of the pattern of abdominal fat distribution includingldquosub-depotsrdquo and its relation to glycemiccontrol and cardiovascular risk in diabeticpatients Effective interventions that canalter abdominal fat distribution may helpsort out whether fat distribution pheno-types are merelya re1047298ection of obesity sub-phenotypes or can causally affect risk andseverity of obesity-associated morbidities

AcknowledgmentsdGrant to IS was pro-vided by the European Foundation for theStudy of Diabetes of the European Associationfor the Study of Diabetes for the study of the 2-year CASCADE randomized controlledtrial

No potential con1047298icts of interest relevant tothis article were reported

RG researched data performed statisticalanalyses and wrote the manuscript IS per-formed data analysis and contributed to dis-cussion AR designed the study performeddata analysis and wrote the manuscript YG

ES YC DS SBA SW IFL OT-Rand BS researched data IH-B and YH re-searched data contributed to discussion andreviewed and edited the manuscript MJSedited the manuscript and contributed to dis-cussion IS (CASCADE Principal Investigator)contributed to discussion and reviewed andedited the manuscript IS is the guarantor of this work and as such had full access to all

the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis

This work is dedicated to Harel Segal ourfellow colleague who died during the study atthe age of 43 years The authors thank the par-ticipants in the CASCADE randomized con-trolled trial for consistent cooperation thefollowing consultants and health care providersHassia Krakauer Meir Aviv Haim StraslerDr Ziva Schwartz Dr Einat Sheiner Dr DovBrickner Rachel Marko Esther Katorza andIlanit Asulin (from the Nuclear Research CenterNegev) Dr Tatiana Shuster Sagit Kachlon Yasmin Asuly and Roman Tsirkin (from the

Soroka University Medical Center) and DrLena Novack and Dana Sarfaty (from TheS Daniel Abraham International Center forHealth and Nutrition Faculty of Health Scien-ces Ben-Gurion University of the Negev)

References1 Hamdy O Porramatikul S Al-Ozairi E

Metabolic obesity the paradox betweenvisceral and subcutaneous fat Curr Dia-betes Rev 20062367ndash373

2 Fox CS Massaro JM Hoffmann U et al Abdominal visceral and subcutaneousadipose tissue compartments associationwith metabolic risk factors in the Fra-mingham Heart Study Circulation 200711639ndash48

3 Matsushita Y Nakagawa T Yamamoto Set al Associations of visceral and sub-cutaneous fat areas with the prevalenceof metabolic risk factor clustering in6292 Japanese individuals the HitachiHealth Study Diabetes Care 2010332117ndash2119

4 Sniderman ADBhopal R Prabhakaran DSarrafzadegan N Tchernof A Why mightSouth Asians be so susceptible to centralobesity and its atherogenic consequences

The adipose tissue over1047298

ow hypothesisInt J Epidemiol 200736220ndash2255 Deschecircnes D Couture P Dupont P

Tchernof A Subdivision of the subcu-taneous adipose tissue compartment andlipid-lipoprotein levels in women ObesRes 200311469ndash476

6 Gallagher D Kelley DEYim JE et al MRI Ancillary Study Group of the Look AHEADResearch Group Adipose tissue distribu-tion is different in type 2 diabetes Am JClin Nutr 200989807ndash814

7 Laakso M Cardiovascular disease in type 2diabetes challenge for treatment and pre-vention J Intern Med 2001249225ndash235

8 Goodpaster BH Thaete FL Simoneau JAKelley DE Subcutaneous abdominalfat and thigh muscle composition pre-dict insulin sensitivity independently of visceral fat Diabetes 1997461579ndash

15859 Porter SA Massaro JM Hoffmann U

Vasan RS OrsquoDonnel CJ Fox CS Abdom-inal subcutaneous adipose tissue a pro-

tective fat depot Diabetes Care 2009321068ndash1075

10 Bays HE Fox KM Grandy S SHIELDStudy Group Anthropometric measure-ments and diabetes mellitus clues to theldquopathogenicrdquo and ldquoprotectiverdquo potential of adipose tissue Metab Syndr Relat Disord20108307ndash315

11 Johnson JA Fried SK Pi-Sunyer FX Albu JB Impaired insulin action in subcuta-neous adipocytes from women with vis-ceral obesity Am J Physiol EndocrinolMetab 2001280E40ndashE49

12 Koska J Stefan N Votruba SB SmithSR Krakoff J Bunt JC Distribution of subcutaneous fat predicts insulin ac-tion in obesity in sex-speci1047297c mannerObesity (Silver Spring) 2008162003ndash

200913 Kelley DE ThaeteFL Troost FHuweT

Goodpaster BH Subdivisions of subcuta-neous abdominal adipose tissue and in-sulin resistance Am J Physiol EndocrinolMetab 2000278E941ndashE948

14 Snijder MB Visser M Dekker JM et alLow subcutaneous thigh fat is a risk factorfor unfavorable glucose and lipid levelsindependently of high abdominal fat TheHealth ABC Study Diabetologia 200548301ndash308

15 Thomas EL Bell JD In1047298uence of under-sampling on magnetic resonance imagingmeasurements of intra-abdominal adi-pose tissue Int J Obes Relat Metab Disord200327211ndash218

16 Task Force of the European Society of Cardiology and the North American So-ciety of Pacing and ElectrophysiologyHeart rate variability standards of mea-surement physiological interpretation andclinical use Circulation 1996931043ndash

106517 Shai I Shahar DR Vardi H Fraser D Se-

lection of food items for inclusion in a

newly developed food-frequency ques-tionnaire Public Health Nutr 20047745ndash749

18 Poanta L Porojan M Dumitrascu DLHeart rate variability and diastolic dys-functionin patientswith type2 diabetesmellitus Acta Diabetol 201148191ndash

19619 M iy az ak i Y G la ss L T ri pl it t C

Wajcberg E Man darino LJ DeF ronzoRA Abdominal fat distribution and pe-ripheral and hepatic insulin resistancein type 2 diabetes mellitus Am J PhysiolEndocrinol Metab 2002283E1135ndash

E1143





20 Bjoumlrntorp P ldquoPortalrdquo adipose tissue as agenerator of risk factors for cardiovasculardisease and diabetes Arteriosclerosis 199010493ndash496

21 Micha R Mozaffarian D Trans fatty acidseffects on cardiometabolic health and im-plications for policy Prostaglandins LeukotEssent Fatty Acids 200879147ndash152

22 Koh-Banerjee P Chu NF Spiegelman D

et al Prospective study of the associationof changes in dietary intake physicalactivity alcohol consumption and smok-ing with 9-y gain in waist circumferenceamong 16 587 US men Am J Clin Nutr200378719ndash727

23 Bertrais S Balkau B Vol S et al Rela-tionships between abdominal body fatdistribution and cardiovascular risk fac-tors an explanation for womenrsquos healthier

cardiovascular risk pro1047297le The DESIRStudy Int J Obes Relat Metab Disord 1999231085ndash1094

24 Rossi AP Fantin F Zamboni GA et alPredictors of ectopic fat accumulation inliver and pancreas in obese men andwomen Obesity (Silver Spring) 2011191747ndash1754

25 Westerbacka J Corneacuter A Tiikkainen M

et alWomen andmen havesimilar amountsof liver and intra-abdominal fat despitemore subcutaneous fat in women impli-cations for sex differences in markers of cardiovascular risk Diabetologia 2004471360ndash1369

26 Picheacute ME Lapointe A Weisnagel SJet al Regional body fat distribution andmetabolic pro1047297le in postmenopausalwomen Metabolism 2008571101ndash1107

27 Taylor HS Manson JE Update in hor-mone therapy use in menopause J ClinEndocrinol Metab 201196255ndash264

28 Harman SM Vittinghoff E Brinton EAet al Timing and duration of menopausalhormone treatment may affect cardiovas-cular outcomes Am J Med 2011124199ndash205

29 Nam JS Nam JY Yoo JS etal The effect of

rosiglitazone on insulin sensitivity andmid-thigh low-density muscle in patientswith Type 2 diabetes Diabet Med 20102730ndash36

30 Virtanen KA Haumlllsten K Parkkola Ret al Differential effects of rosiglitazoneand metformin on adipose tissue dis-tribution and glucose uptake in type 2diabetic subjects Diabetes 200352283ndash

290





an insulin pump or injecting insulin morethan twice per day had evidence of severediabetes complications (eg proliferativeretinopathy or advanced renal disease)had autonomic neuropathy manifested aspostural hypertension or hypoglycemiaunawareness had a fasting serum triglyc-eride level 400 mgdL or a serum creat-

inine level of $2 mgdL (177 mmolL)had liver dysfunction (greater than twicethe upper limit of normal of alanine ami-notransferase and aspartate aminotransfer-ase levels) had active cancer or receivedchemotherapy in the last 3 years or wereparticipating in another trial The study protocol was approved by the Soroka Uni-versity Medical Center Medical EthicsBoard and Helsinki Committee All partic-ipants provided written informed consent

MRI acquisition and image analysisMRI scans of the abdomen were performedusing a 15 Tesla machine (Intera PhilipsMedical Systems Best the Netherlands)using a body coil Subjects were examinedin thesupineposition witharms positionedparallel along the lateral sides of the bodyMRI scans demonstrating fat in the differ-ent compartments were assessed using aMATLAB-based in-house program (Fig 1)Fat tissues of speci1047297c anatomical land-marks were quanti1047297ed The MRI scan al-lows visualizing the fascia super1047297cialisas a 1047297ne black line To divide SSAT fromDSAT we drew a continuous line over the

fascia super1047297cialis After quanti1047297cation fattissues were divided into color-codedgroups super1047297cial subcutaneous fat =dark blue deep subcutaneous fat = lightblue VAT = green perimuscular fat (fatsurrounding and within the latissimusdorsi and diaphragm) = purple and non-classi1047297ed fat (fat surrounding the vertebraeand fat depots unrelated to any of thegroups listed above) = red Selecting thespeci1047297ed fat mass area was performedusing a semiautomatic method (ie con-nected pixels) or various manual toolssuch as rectanglecirclepolygon or freehand for 1047297ne adjustments and correctionsif needed Quanti1047297cation of the fat massregions included the area of each fat typeand the proportion(percentage) of the totalarea of all fat types To obtain absolutemeasurements in metric units a scalingprocedure was applied before the segmen-tation to determine real pixel dimensionsFinally in accordance with other studies(15) we calculated the fat distribution us-ing the mean of the three slices S1L5L5L4 andL3L2Perimuscular and thenon-classi1047297ed fat tissues totaling a negligible

fraction of total abdominal fat were omittedfrom our analysis All multivariate analyseswere performed twice using the absoluteabdominal fat tissue distribution (SSATDSAT VAT) or the relative (in percent) of each depot from total abdominal adiposetissue (TAAT) (SSAT DSAT VAT)

Clinical parameters Anthropometric measures were evaluated

as clinical estimates of whole-body adi-pose tissue compartments Participantswere weighed without shoes to the nearest01 kg Height was measured using a wall-mounted stadiometer to the nearest mil-limeter for determination of BMI Waistcircumference was measured halfway be-tween the last rib and the iliac crest by thequali1047297ed study nurse with the same typeof measuring tape Mean blood pressurefrom two measures was recorded afterresting with the use of an automated sys-tem (Datascop Acutorr 4 SOMA Technol-ogy Inc Bloom1047297eld CT) Blood sampleswere drawn after at least an 8-h fast andcurrent use of all medication was recordedTo assess blood pressure and heart ratevariability 24-h ambulatory blood pres-sure monitoring (OSCAR 2 oscillometricSunTech Medical Model 222 MorrisvilleNC) and 24-h ambulatory electrocardi-ography (ECG) (LIFECARD-CF DelmarReynolds Medical Ltd Hertfordshire UK)were performed in a substudy group of ourpopulation (n = 31 and n = 37 respec-tively) Heart ratevariability parameters in-cluded time domain variables (16) the

mean duration of the timeinterval betweentwo R waves (RR) graphically presented inthe form of an RR interval tachogram theSD of all normal RR (SDNN) the mean of all the 5-min SDs of NN (normal RR) inter-vals during the 24-h period the root meansquare successive difference calculatesthe square root of the mean of the squareddifferences between successive NN inter-vals over 24 h and the 24-h triangular in-dex the integral of the density distribution

(ie the number of all NN intervals) di-vided by the maximum of the density dis-tribution which is more in1047298uenced by thelower than the higher frequencies Fromthe 24-h blood pressure monitor we calcu-lated the average systolic and diastolicblood pressure during the day (average of records between 6 AM and 10 PM) andnight (average of records between 10 PM

and 6 AM) All medications in current usewere recorded and classi1047297ed

Dietary assessment We evaluated dietary assessment by avalidated food-frequency questionnaire(17) that included 127 food items and 3portion-size pictures for17 items Electronicquestionnaires ensured completeness of thedata by prompting the participant when aquestion was not answered

Statistical analysis We divided our study population acrosstertiles of SSAT (range 5052ndash26836mm2) and used both absolute and pro-portional fat distribution for the analysis

Figure 1dMRI imaging of abdominal fat tissues compartments The subcutaneous fascial planewas delineated using the computer interface semiautomatic method where initially an intensity-based automatic segmentation was generated and presented followed by semimanual 1047297ne tuningThe area of each compartment was quantitated separately Fat tissues of speci 1047297c anatomicallandmarks were quanti 1047297ed and divided into color-coded groups as follows dark blue super 1047297cialsubcutaneous fat light blue deep subcutaneous fat green visceral adipose tissue and red non-

classi 1047297

ed fatd

fat surrounding the vertebrae and fat depots that were unrelated to each of the groupslisted above (A high-quality digital representation of this 1047297gure is available in the online issue)





T a b l e





E n t i r e

P

o f t r e n d

L o w t e r t i l e


T

o p t e r t i l e

R a n g e 5 0 5 2 ndash 7 2 3 1 m m

2

R a n g e 7 2 8 2 ndash 9 8 9 8 m m

2

R a n g e 9 9 3 0 ndash 2 6 8 3 6 m m

2

R a n g e 5 0 5 2 ndash 2 6 8

3 6 m m

2

S S A T m m

2

6 4 0 9 6

6 2 5

8 5 2 5 6

7 6 9

1 3 4

5 6 6

3 7 9 0

9 4 6 0 6

3 7

7 7

d

A g e y e a r s

5 8 7 6

8 8

5 8 5 6

5 7

5 8 4 6

7 9

5 8 8 6

7 4

N S

M a l e s e x n

( )

2 6 ( 1 0 0 )

1 9 ( 8 6 )

1 6 ( 6 4 )

6 1 ( 8 3 )

0 0 0 1

W e i g h t k g

8 5 2 6

1 1 7

8 6 3 6

1 4 2

8 8 5 6

1 4 1

8 6 6 6

1 3 3

N S


1 0 1 1 6

7 1

1 0 3 7 6

1 0 3

1 0 6 5 6

9 2

1 0 3 8 6

9 2

0 0 1 4

B M I k g m

2

2 8 0 6

3 5

2 8 9 6

3 7

3 1 0 6

3 8

2 9 2 6

3 8

0 0 0 1


1 3 8 4 6

1 6 0

1 3 1 1 6

1 4 4

1 3 9 7 6

1 8 3

1 3 6 2 6

1 6 4

N S


8 0 5 2 6

1 0 2

7 8 4 6

9 3

7 8 1 6

1 1 1

7 9 0 6

1 0 1

N S



2

1 2 7 3 3 6

2 9 2 9

1 7 4 6 7 6

3 3 2 5

2 4 0

1 8 6

4 6 5 4

1 8 0 2 4 6

6 0

0 8

0 0 0 1


2

6 3 2 4 6

2 6 1

8 9 4 1 6

3 1 1 0

1 0 5

6 1 6

3 1 6 9

8 5 6 4 6

3 4

2 9

0 0 0 1

V A T m m

2

1 8 7 8 0 6

5 4 6 0

2 1 4 5 5 6

7 0 5 6

1 7 3

1 4 6

6 8 4 6

1 9 0 8 4 6

6 5

8 3

N S

T A A T m m

2

3 1 5 1 4 6

6 0 0 4

3 8 9 2 2 6

8 6 1 7

4 1 3

3 2 6

7 8 8 0

3 7 1 0 9 6

8 5

6 6

0 0 0 1



o f T A A T

2 0 8

2 2 7

3 3 2

2 5 6

d



5 2 4

5 0 0

5 6 0

5 2 9

d


0 3 6 6

0 1

0 4 3 6

0 1

0 8 9 6

0 4

0 5 6 6

0 3

d


F P G m g d L

1 6 1 1 6

7 5 0

1 4 2 9 6

2 9 7

1 2 7 8 6

2 3 0

1 4 3 7 6

4 8 9

0 0 4 4

H b A 1 c

7 9 6

1 4

7 5 6

1 0

7 0 6

0 4

7 5 6

1 1

9

0 1 0


1 3 6 5 6

6 3 0

1 7 9 5 6

8 4 8

1 6 6 0 6

8 3 6

1 6 0 8 6

7 8 8

N S


m g d L

4 5 0 6

6 6

4 1 6 6

1 1 6

4 8 9 6

1 0 2

4 5 1 6

1 0 0

N S


9 2 7 6

3 1 4

9 7 4 6

3 0 5

1 0 1 6 6

2 7 8

9 6 8 6

2 9 9

N S


7 0 4 6

2 5 1

8 1 0 6

3 4 8

6 3 1 6

1 5 4

7 2 3 6

2 7 8

N S


2 5 2 6

1 0 0

2 8 9 6

1 1 1

2 6 2 6

1 1 8

2 6 8 6

1 0 9

N S



( )

3 ( 1 1 )

5 ( 2 3 )

2 ( 8 )

1 0 ( 1 4 )

N S


( )

1 7 ( 6 5 )

1 8 ( 8 2 )

1 6 ( 6 4 )

5 1 ( 7 0 )

N S


( )

1 3 ( 5 0 )

1 5 ( 6 8 )

1 6 ( 6 4 )

4 4 ( 6 0 )

N S


( )

1 3 ( 5 0 )

1 4 ( 6 3 )

1 4 ( 5 6 )

4 1 ( 5 6 )

N S


( )

1 2 ( 4 6 )

1 1 ( 5 0 )

1 2 ( 4 8 )

3 5 ( 4 8 )

N S




P o

f t r e n d












RESULTS


























Men Women

Age years 584 6 75 597 6 72

Weight kg 877 6 131 815 6 134

BMI kgm2 289 6 37 312 6 38



Abdominal MRI

Total SAT mm2 17080 6 5312 22825 6 7230



VAT mm2 20014 6 6629 14355 6 3838

TAAT mm2 37095 6 8848 37180 6 7290

Fat distributions







HbA1c 76 6 11 69 6 04

Serum TG mgdL 1622 6 832 1540 6 540



TGHDL 38 6 23 34 6 19















































































E1143




















290





T a b l e





E n t i r e

P

o f t r e n d

L o w t e r t i l e


T

o p t e r t i l e

R a n g e 5 0 5 2 ndash 7 2 3 1 m m

2

R a n g e 7 2 8 2 ndash 9 8 9 8 m m

2

R a n g e 9 9 3 0 ndash 2 6 8 3 6 m m

2

R a n g e 5 0 5 2 ndash 2 6 8

3 6 m m

2

S S A T m m

2

6 4 0 9 6

6 2 5

8 5 2 5 6

7 6 9

1 3 4

5 6 6

3 7 9 0

9 4 6 0 6

3 7

7 7

d

A g e y e a r s

5 8 7 6

8 8

5 8 5 6

5 7

5 8 4 6

7 9

5 8 8 6

7 4

N S

M a l e s e x n

( )

2 6 ( 1 0 0 )

1 9 ( 8 6 )

1 6 ( 6 4 )

6 1 ( 8 3 )

0 0 0 1

W e i g h t k g

8 5 2 6

1 1 7

8 6 3 6

1 4 2

8 8 5 6

1 4 1

8 6 6 6

1 3 3

N S


1 0 1 1 6

7 1

1 0 3 7 6

1 0 3

1 0 6 5 6

9 2

1 0 3 8 6

9 2

0 0 1 4

B M I k g m

2

2 8 0 6

3 5

2 8 9 6

3 7

3 1 0 6

3 8

2 9 2 6

3 8

0 0 0 1


1 3 8 4 6

1 6 0

1 3 1 1 6

1 4 4

1 3 9 7 6

1 8 3

1 3 6 2 6

1 6 4

N S


8 0 5 2 6

1 0 2

7 8 4 6

9 3

7 8 1 6

1 1 1

7 9 0 6

1 0 1

N S



2

1 2 7 3 3 6

2 9 2 9

1 7 4 6 7 6

3 3 2 5

2 4 0

1 8 6

4 6 5 4

1 8 0 2 4 6

6 0

0 8

0 0 0 1


2

6 3 2 4 6

2 6 1

8 9 4 1 6

3 1 1 0

1 0 5

6 1 6

3 1 6 9

8 5 6 4 6

3 4

2 9

0 0 0 1

V A T m m

2

1 8 7 8 0 6

5 4 6 0

2 1 4 5 5 6

7 0 5 6

1 7 3

1 4 6

6 8 4 6

1 9 0 8 4 6

6 5

8 3

N S

T A A T m m

2

3 1 5 1 4 6

6 0 0 4

3 8 9 2 2 6

8 6 1 7

4 1 3

3 2 6

7 8 8 0

3 7 1 0 9 6

8 5

6 6

0 0 0 1



o f T A A T

2 0 8

2 2 7

3 3 2

2 5 6

d



5 2 4

5 0 0

5 6 0

5 2 9

d


0 3 6 6

0 1

0 4 3 6

0 1

0 8 9 6

0 4

0 5 6 6

0 3

d


F P G m g d L

1 6 1 1 6

7 5 0

1 4 2 9 6

2 9 7

1 2 7 8 6

2 3 0

1 4 3 7 6

4 8 9

0 0 4 4

H b A 1 c

7 9 6

1 4

7 5 6

1 0

7 0 6

0 4

7 5 6

1 1

9

0 1 0


1 3 6 5 6

6 3 0

1 7 9 5 6

8 4 8

1 6 6 0 6

8 3 6

1 6 0 8 6

7 8 8

N S


m g d L

4 5 0 6

6 6

4 1 6 6

1 1 6

4 8 9 6

1 0 2

4 5 1 6

1 0 0

N S


9 2 7 6

3 1 4

9 7 4 6

3 0 5

1 0 1 6 6

2 7 8

9 6 8 6

2 9 9

N S


7 0 4 6

2 5 1

8 1 0 6

3 4 8

6 3 1 6

1 5 4

7 2 3 6

2 7 8

N S


2 5 2 6

1 0 0

2 8 9 6

1 1 1

2 6 2 6

1 1 8

2 6 8 6

1 0 9

N S



( )

3 ( 1 1 )

5 ( 2 3 )

2 ( 8 )

1 0 ( 1 4 )

N S


( )

1 7 ( 6 5 )

1 8 ( 8 2 )

1 6 ( 6 4 )

5 1 ( 7 0 )

N S


( )

1 3 ( 5 0 )

1 5 ( 6 8 )

1 6 ( 6 4 )

4 4 ( 6 0 )

N S


( )

1 3 ( 5 0 )

1 4 ( 6 3 )

1 4 ( 5 6 )

4 1 ( 5 6 )

N S


( )

1 2 ( 4 6 )

1 1 ( 5 0 )

1 2 ( 4 8 )

3 5 ( 4 8 )

N S




P o

f t r e n d












RESULTS


























Men Women

Age years 584 6 75 597 6 72

Weight kg 877 6 131 815 6 134

BMI kgm2 289 6 37 312 6 38



Abdominal MRI

Total SAT mm2 17080 6 5312 22825 6 7230



VAT mm2 20014 6 6629 14355 6 3838

TAAT mm2 37095 6 8848 37180 6 7290

Fat distributions







HbA1c 76 6 11 69 6 04

Serum TG mgdL 1622 6 832 1540 6 540



TGHDL 38 6 23 34 6 19















































































E1143




















290








RESULTS


























Men Women

Age years 584 6 75 597 6 72

Weight kg 877 6 131 815 6 134

BMI kgm2 289 6 37 312 6 38



Abdominal MRI

Total SAT mm2 17080 6 5312 22825 6 7230



VAT mm2 20014 6 6629 14355 6 3838

TAAT mm2 37095 6 8848 37180 6 7290

Fat distributions







HbA1c 76 6 11 69 6 04

Serum TG mgdL 1622 6 832 1540 6 540



TGHDL 38 6 23 34 6 19















































































E1143




















290














Men Women

Age years 584 6 75 597 6 72

Weight kg 877 6 131 815 6 134

BMI kgm2 289 6 37 312 6 38



Abdominal MRI

Total SAT mm2 17080 6 5312 22825 6 7230



VAT mm2 20014 6 6629 14355 6 3838

TAAT mm2 37095 6 8848 37180 6 7290

Fat distributions







HbA1c 76 6 11 69 6 04

Serum TG mgdL 1622 6 832 1540 6 540



TGHDL 38 6 23 34 6 19















































































E1143




















290






































































E1143




















290













































E1143




















290




















290



Documents

Abdominal Superficial Subcutaneous Fat