You are on page 1of 32

Clinical In-Service

:
Using Ultrasound and Computed
Tomography as a Measure of Body
Composition
Aµrll ç
th
, zc¡µ
Luryssu Crgurlc
The Need For Body Composition Methods
• 8ody Comµosltlon ls un
lndlcutor of nutrltlonul
stutus
• Eelµs muke ussessments of
µutlents stronger:
– An uthlete muy huve skewed
8Ml due to ! L8M
– Alternutlvely, un elderly
µerson muy huve " bone
denslty & L8M und muy
welgh less thun u younger
udult of sume stuture
– Cbese or overwelght
lndlvlduuls muy huve !
udlµose tlssue but " L8M
– Llñerences between ruces
• Aglng cuuses " muscle und ! ln fut muss
– Colned term: !"#$%&'()$ %+'!),-
– lncreuslng µrevulence ln surcoµenlc obeslty ln the elderly
– Surcoµenlu " functlonullty, ! rlsk of fullsjfx, ! hosµltul stuys,
! noscomlul lnfectlons & ls ussocluted wlth µoor outcomes
ln nonmullgnunt dlseuses
• Cuchexlu redefned:
– "Cuchexlu ls u comµlex metubollc syndrome ussocluted wlth
underlylng lllness und churucterlzed by loss of muscle !"#$
&' !"#$&(# )&** &+ +,# -,**. 1he µromlnent cllnlcul feuture of
cuchexlu ls welght loss ln udults (corrected for fuld
retentlon) or growth fullure ln chlldren.'
• Muscle loss ln the ubsence of chunges ln fut muss.

!"#$%&'(&()$ +"$ )++%' "& "$, -!./
• luwukuml et ul. found ufter zc duys, muscle
muss ln LL " muscle muss by ¸µ° of strlct
bed rest ln .'"/,.- $%(,#%/!

• Reld et ul. showed " of muscle thlckness ln
$#),)$"//- )// &",)'(,! by ¡.6° µer duy
Methods of Measuring Body Composition
01($ 2)#3 !4"5%'464$&5
• Accurucy deµends on technlque
und reµetltlon
• Assumes ~¸c° of body fut ls
subcutuneous
• Accurucy decreuses wlth
lncreused obeslty
• Stundurds of meusurement ure
bused on '(78& slde of the body
– 1rlceµs
– 8lceµs
– Abdomlnul
– ChestjPectorul
– Medlul Culf
– Mlduxlllury
– Subscuµulur
– Suµrullluc
– 1hlgh
• *SLL = ± ¸.¸°
1rlceµs Skln lold
*Stundurd Lrror of the Lstlmute
Methods of Measuring Body Composition
9('+%6:4'4$+4 !4"5%'464$&5
• vulsttohlµ clrcumference rutlo
– Llµodystroµhy (Elv)
– Curdlovusculur rlsk
– Rlsk: c.8÷ ln ♀, ¡÷ ln ♂
• vulst Clrcumference
– >µc' ♂, >¸¸' ln ♀
– Not useful for those >6c' tull or
wlth 8Ml >¸¸
– Elghly correluted wlth vlscerul
udlµose tlssue
• Mldurm Clrcumference (MAC)
– Eulfwuy µolnt between ucromlon
µrocess & olecrunon µrocess
– 8onefree Arm Muscle Areu
(AMA): ¡c for ♂, 6.¸ for ♀
• Eeud Clrcumference
– Chlldren >¸ yeurs
– lndlcutor of nonnutrltlonul
ubnormulltles
vulst Clrcumference
Methods of Measuring Body Composition
;(' <(5=#"+464$& >#4&8,56)7'"6
• Meusurement of body
denslty to estlmute body
fut und futfree mussees
• Accurute, relluble
• Loes not rely on body E
z
c
content - useful ln LSRL
• Llmltutlons:
– 8udget
– Portublllty
– Cluustroµhoblu?
– 8ed bound µutlent?
• SLL = ± z., - ¸.,°
"8odPod'
Methods of Measuring Body Composition
-()4#4+&'(+"# .6=43"$+4 ;$"#,5(5
• 8used on theory thut vs. wuter
tlssue us u hlgher electrlcul
conductlvlty und lower
lmµedunce vs. udlµose tlssue
becuuse of electrolyte content
• Sufe, nonlnvuslve, µortuble,
ruµld
• Accurucy uñected by hydrutlon
stutus, umong other fuctors:
– Pt should be well hydruted
– No consumµtlon of L1CE,
cuñelne or dluretlcs ln µust zµ
hours
– No exerclse (~µ6 hrs µrevlous)
• Some methods muy not be us
uccurute (ln communlty settlng)
• Loes not dlstlngulsh skeletul
muscle from other soft leun
tlssue
• SLL = ± ¸.¸°
8lA
Methods of Measuring Body Composition
<%"# ?$4'7, @AB", ;C5)'=&()64&', D<?@;E
• Assesses bone mlnerul denslty but
ulso meusures fut und L8M
• ×ruy tube thut contulns energy
beum
– Lnergy loss deµends on tyµe of tlssue
whlch the beum µusses through
• Low rudlutlon
• Reudlly uvulluble ln most cllnlcul
settlngs
• Añected by hydrutlon stutus &
bonejculclfed soft tlssue cun result
ln lnuccurucles
• Llsudvuntuges: cost, not µortuble &
lnvuslve
• /&)&0"1 2"*1&34'5 model used most
commonly ln revlewed urtlcles
• SLL = ± ¡.8°
LL×A
Methods of Measuring Body Composition
F,3')5&"&(+ G4(78($7
– Llrect meusure bused on
Archlmedes' Prlnclµle
• volume of un object submerged ln
E
z
c equuls the volume of E
z
c the
object dlsµluces
• vhen volume und Muss ls known,
denslty ls culculuted
– Consldered the "Cold Stundurd'
for body comµosltlon
meusurements but does not
dlñerentlute between musclejfut
tlssues
– Llmcult for those meusured us
lndlvlduul must remuln under
wuter for extended µerlods of tlme
– Person µerformlng the test must
be trulned
– lmµructlcul ln most settlngs other
thun reseurch
– SLL = ± z.,°
Other Measures of Body Composition
• H4%&')$ ;+&(I"&()$ ;$"#,5(5
– Meusures L8M
– Creutes unstuble lsotoµes of Cu, N & Nu then meusures
gummu rudlutlon
– Lxµenslve, lmµructlcul ln most settlngs
• J)&"# -)3, K
– >çc° of body l ls found ln fut free tlssue
– Uses gummu ruy detectors lnterfuced wlth comµuter
– Lxµenslve, not reudlly uvulluble. Reseurchers do not ugree on
l concentrutlon between dlñerent body tyµes e.g. elderly,
obese lndlvlduuls
• !"7$4&(+ B45)$"$+4 .6"7($7
– Accurutely meusures overull body comµosltlon but cun
overestlmute fut deµoslts
– Lxµenslve, long wult tlmes
Computed Tomographic (CT) Scan
• Consldered the gold stundurd for "3(=)54 &(55%4
meusurement
– Refects wlde vurlublllty wlthln 8Ml
• Eounsfeld Unlt (EU)
– Pructlcul threshold unlt used ln C1 - unlts of
rudlutlon uttenuutlon or "lntenslty'
• zç to ¡¸c EU for skeletul muscle
• ¡çc to ¸c EU subcutuneous und lntermusculur udlµose
tlssue
• ¡¸c to ¸c EU vlscerul udlµose tlssue
• Requlres u llcensed rudlogruµher (¸yeur
dlµlomu) wlth extenslve exµerlence
CT Continued
• L¸ Lumbur vertebrul Lundmurk
– Skeletul muscle & udlµose tlssue
$%##'!&%(0 ,% 1.%/'2+%0- ,)!!3'
43"(,),)'! - most common slte
used ln revlewed urtlcles
• Contulns vlscerul und
subcutuneous udlµose tlssue
• Purusµlnul muscles, trunsversus
ubdomlnus, externul und lnternul
ollque ubdomlnuls und rectus
ubdomlnus
• Lµ Lumbur vertebrul Lundmurk
ulso used
– Above the llluc crest
– 'ustlfcutlon: to uvold µosslble
lnterference from llluc rest bone
µlxels.
– Aµµroµrlute when comµurlng C1
to LL×A.
Lµ Lumbur vertebrul Lundmurk
Structures Identified in CT Scan
Baracos
23
Experts are starting to acknowledge the independent behavior of muscle and
adipose tissues on wasting syndromes. A recently convened consensus conference
on the definition of cachexia
13
notably made a distinction between the behavior of
skeletal muscle and adipose tissue: “Cachexia, is a complex metabolic syndrome
associated with underlying illness and characterized by loss of muscle with or
without loss of fat mass. The prominent clinical feature of cachexia is weight loss in
adults (corrected for fluid retention) or growth failure in children….” This perspective
acknowledges the possibility of a persistent muscle loss in the absence of any
change in fat mass.
The Use of Diagnostic Images To Assess Body
Composition Changes and Sarcopenia in Cancer Patients
A highly differentiated understanding of human body composition has evolved
in tandem with image-based technologies such as computed tomography (CT)
and magnetic resonance imaging (MRI). These methods enjoy a high degree of
specificity for the separate discrimination of many organs and tissues (Fig 1).
Pancreas
Duodenum
Ascending colon
Liver
External abdominal oblique
Transversus abdominus
Internal abdominal oblique
Right kidney
Inferior vena cava Subcutaneous adipose
Visceral adipose
Left kidney
Abdominal aorta
Descending colon
Jejunum
Ileum
Superior mesenteric a.
Superior mesenteric v.
Rectus abdominus Linea alba Transverse colon
Quadratus lumborum
Erector spinae muscles Intramuscular adipose tissue
Psoas major and minor Third lumbar vertebra
Fig 1. Structures present in a computerized tomography image at the 3rd lumbar vertebra.
vlckle 8urucos, PhL
Volume Calculations
• Used by lvlst et ul (¡ç88)
to culculute totul vlscerul
volume by uslng totul
vlscerul ureus und
dlstunces between the
scuns
• Alm of the study wus to
fnd u better method of
comµosltlon vs. C1
becuuse of lts exµense.
• SLL between
unthroµometry und C1
regurdlng suglttul
dlumeters were ±¡.8° ut
L¸Lµ und ±¡.,° LµL¸
Image Analysis Software
• 8onekumµ et ul, (zcc,): Study µubllshed ln the
lnternutlonul 'ournul of Cbeslty
• 1here ure ¸ µrogrum uvulluble to ussess ubdomlnul
udlµose tlssue:
– NlE lmuge' - free
– Sllceomutlc
– Anulyze
– Elµµofut - free
– Lusyvlslon - used wlth Phllllµs muchlnes
• lound lt tukes ¸z seconds to segment u scun uslng
Sllceomutlc, ¸zz.¸ seconds (u llttle over ¸ mlnutes) to use
lmuge'.
• Results uslng SllceCmutlc, Anulyze or NlE lmuge' were
comµuruble und could be used lnterchungeubly
• Munuul, semluutomuted or uutomuted segmentutlon


Analysis of CT Scan
Baracos
25
Cross-sectional areas (cm
2
) are computed automatically by the program once
the desired tissues are demarcated. Demarcated images (Fig 2) illustrate body
composition changes in a lung cancer patient over time; this person lost skeletal
muscle during progressive disease.
1 2
Fig 2. Skeletal muscle loss in a patient with lung cancer. Segmented CT images for a male
lung cancer patient at two separate timepoints. Number 1 was taken 390 days before death,
and 2 was taken 58 days before death. Segmented tissues of interest: is skeletal muscle,
is visceral adipose tissue, is subcutaneous adipose tissue, and is intramuscular adipose
tissue. During this 332-day period, skeletal muscle area decreased from 173 cm
2
to 86.7 cm
2
.
Conclusions
Current demographics of weight and body composition suggest a need to
reconceptualize cancer cachexia. Substantial depletion of skeletal muscle is a
widespread abnormality of body composition in patients with advanced solid
tumors, which is present in people at any BMI and strongly related to outcome. Valid
and convenient approaches for determining muscularity are required to evaluate this
feature in cancer patients, and the secondary analysis of CT images is an accessible
means of making this evaluation.
References
1. Yavuzsen T, Davis MP, Walsh D, et al: Systematic review of the treatment of
cancer-associated anorexia and weight loss. J Clin Oncol 2005;23:8500-8511.
2. Prado CM, Lieffers JR, McCargar LJ, et al: Prevalence and clinical implications
of sarcopenic obesity in patients with solid tumours of the respiratory and
gastrointestinal tracts: A population-based study. Lancet Oncol 2008;9:629-635.
3. Baumgartner RN, Koehler KM, Gallagher D, et al: Epidemiology of sarcopenia
among the elderly in New Mexico. Am J Epidemiol 1998;147:755-763.
1lssues ure ldentlfed by thelr unutomlcul feuture und then segmented und quuntlfed bused on
Eounsfeld unlts uslng lmuglng unulysls softwure.
Crosssectlonul ureus ln cm
z
ure comµuted uutomutlcully by the softwure once deslred tlssues
ure segmented.
1hls lmuge shows body comµosltlon ln u lung cuncer µutlent over tlme. #¡: ¸çc duys before
deuth. #z: ¸8 duys before deuth. As the dlseuse µrogressed, skeletul muscle decreused from
¡,¸ cm
z
to 86., cm
z

Skeletul Muscle
vlscerul Adlµose 1lssue
Subcutuneous Adlµose 1lssue lntrumusculur Adlµose 1lssue
vlckle 8urucos, PhL
Tomovision Sliceomatic
• 634'3"4! +'&- #$4 !47*"#48
– SllceCmutlc wus develoµed for reseurchers ln 8ody Comµosltlon. lts
evolutlon ls bused on yeurs of colluborutlon between the
µrogrummers und the reseurchers who use lt.

– vlth SllceCmutlc, morµhologlcul oµerutors cun eñortlessly µerform
tlssue segmentutlon. lurthermore, studles huve µroven thut thls
technlque ls generutlng conslstent results.

– Llñerent modulltles or even dlñerent tlssues wlthln the sume sllce
ure best segmented wlth dlñerent technlques. SllceCmutlc mukes lt
eusy for you to use the uµµroµrlute technlque for euch sµeclfc tusk.
lor exumµle, ln un MRl sllce of the ubdomen, you muy wunt to use
muthemutlcul morµhology to segment the subcutuneous fut und
reglongrowlng to segment the lntruubdomlnul fut.

• Cost: sµccc.cc
Ultrasound (US)
• US most uccurute when
comµured to cullµers und
hydrostutlc welghlng & US
guve best µredlctlon of
body denslty ut wulst slte
(r=c.,¸6, µ<c.ccc¡) ln leun,
heulthy subjects
(Am ' Clln Nutr ¡ç8µ,¸ç:,c¸,cç)
• Posltlon, locutlon,
orlentutlon to slte und
contuct µressure ull fuctors
ln uccurucy - requlres
hlghly skllled technlclun
• Sltes used: ubdomlnul, mld
thlgh, umblllcul level
2 ISRN Obesity
FicUvi I: Ultrasound measurement points at the posterior abdominal wall and midthigh (right and lef).
time. Because of statistically signifcant diferences between
measurements obtained with diferent machines [I0–I3],
users had to carry out new cross-validation studies, with a
view to longitudinal followup studies [I4–I6].
All these reasons lead us to question about the perfor-
mance of the portable ultrasound-based device, in which
measurement of total body fat mass had been estimated in
2002 by comparison with data from a Hologic QDR-4300W
DXA instrument.
According to the fact that total body fat obtained by
two diferent DXA has a signifcant individual diference,
we consider that DXA absportiometry techniques are not
interchangeable. For all these reasons, we were led to envisage
a newcalibrationof our portable ultrasounddevice using data
collected with a more contemporary DXA instrument.
2. Subjects and Methods
Data was collected on a sample consisted of 83 sedentary
women and 4I sedentary men aged respectively 40.1 ± 15.1
and 48 ± 15.3 years at Nancy University Hospital. Only
patients who gave written consent afer receiving a letter of
information on the measurement protocol were included. All
patients in good health were recruited according to a wide
range of body mass index (BMI) and consequently of total
fat mass. In each subject, body composition was measured by
DXA radiological examination and by our ultrasound-based
device on the same day. All subjects had eaten breakfast in the
morning and were properly hydrated before measurement.
Te reference measures of total body fat mass (BF in kg)
were obtained by a Hologic Discovery A(version I2.7.2). Tis
technique, which scans the whole body with an X-ray beamat
two energy levels (40 and I00 Kev), is a reference method for
measurement of fat mass, leanmass, andmineral content. Te
subject lies in a supine position for 7 minutes and radiation
exposure is very low. Te subject’s weight is calculated with
an accuracy of less than I%. Te results of body composition
are available immediately afer each examination.
Ultrasound measurements were made with a sonographic
US BOX in A-mode from Lecoeur Electronique Company
(Chuelles, France). Absorption of ultrasound depends on
probe depth as well as on the square of the frequency of
the waves transmitted. Te US can be used to measure
the thickness of subcutaneous fat between the skin and the
muscle. We selected two preferred anatomical areas: the
abdominal areas which are ofen associated with metabolic
risk factors [I7–I9] and the midthigh area. Subcutaneous
fat was located in a horizontal plane with approximately
43

axis vertebral at umbilical level and also at the middle
of the knee and the top thigh on the anterior side with a
2.23 MHz linear probe (Figure I). We used a probe with a
0.73 inches diameter that is the most appropriate in terms
of positioning, location, orientation, and contact pressure.
Interobserver reproducibility of fat thickness measurements
with the ultrasound technique was good, and intraclass
correlation obtained by two examiners in the same subjects
was greater than 0.98.
Anthropometric measurements, weight, height, and,
umbilical waist circumference, were recorded by the same
operator using standard anthropometric techniques [20].
.. Statistics. Fat mass estimation by ultrasound uses a linear
regression equation developed from DXA reference value.
Tis regressionequationincludes the anthropometric charac-
teristics and subcutaneous fat thickness at the midthigh and
the back of the umbilical level. Statistical analysis was carried
out to compare the fat mass obtained by the ultrasound
technique and by DXAabsorptiometry using Student’s paired
?-test. Te accuracy of the fat mass estimations calculated
through the preceding models was evaluated from the deter-
mination coemcient ?
2
between true and estimated values
of fat mass and the standard error of the estimate (SEE)
as described by Lohman [2I]. Agreement between body
composition estimates was examined by calculating the 93%
limits of agreement as described by Bland and Altman [22].
Additionally, potential bias between BF estimates by DXA
and the US technique was obtained using residual plots. For
all analysis ? < 0.05 was considered signifcant. Statistical
tests were performed using Statistica sofware (version 6;
StatSof, Tulsa, Okla., USA).
3. Results
Te descriptive characteristics of the anthropometric mea-
surements, fat thicknesses measured by ultrasound and the
total fat mass (BF in kg) and BF% by DXA of sedentary
women and men separately are given in Table I.
Te regression equations used to estimate total fat mass
according to Hologic Discovery A are as follows.
2 ISRN Obesity
FicUvi I: Ultrasound measurement points at the posterior abdominal wall and midthigh (right and lef).
time. Because of statistically signifcant diferences between
measurements obtained with diferent machines [I0–I3],
users had to carry out new cross-validation studies, with a
view to longitudinal followup studies [I4–I6].
All these reasons lead us to question about the perfor-
mance of the portable ultrasound-based device, in which
measurement of total body fat mass had been estimated in
2002 by comparison with data from a Hologic QDR-4300W
DXA instrument.
According to the fact that total body fat obtained by
two diferent DXA has a signifcant individual diference,
we consider that DXA absportiometry techniques are not
interchangeable. For all these reasons, we were led to envisage
a newcalibrationof our portable ultrasounddevice using data
collected with a more contemporary DXA instrument.
2. Subjects and Methods
Data was collected on a sample consisted of 83 sedentary
women and 4I sedentary men aged respectively 40.1 ± 15.1
and 48 ± 15.3 years at Nancy University Hospital. Only
patients who gave written consent afer receiving a letter of
information on the measurement protocol were included. All
patients in good health were recruited according to a wide
range of body mass index (BMI) and consequently of total
fat mass. In each subject, body composition was measured by
DXA radiological examination and by our ultrasound-based
device on the same day. All subjects had eaten breakfast in the
morning and were properly hydrated before measurement.
Te reference measures of total body fat mass (BF in kg)
were obtained by a Hologic Discovery A(version I2.7.2). Tis
technique, which scans the whole body with an X-ray beamat
two energy levels (40 and I00 Kev), is a reference method for
measurement of fat mass, leanmass, andmineral content. Te
subject lies in a supine position for 7 minutes and radiation
exposure is very low. Te subject’s weight is calculated with
an accuracy of less than I%. Te results of body composition
are available immediately afer each examination.
Ultrasound measurements were made with a sonographic
US BOX in A-mode from Lecoeur Electronique Company
(Chuelles, France). Absorption of ultrasound depends on
probe depth as well as on the square of the frequency of
the waves transmitted. Te US can be used to measure
the thickness of subcutaneous fat between the skin and the
muscle. We selected two preferred anatomical areas: the
abdominal areas which are ofen associated with metabolic
risk factors [I7–I9] and the midthigh area. Subcutaneous
fat was located in a horizontal plane with approximately
43

axis vertebral at umbilical level and also at the middle
of the knee and the top thigh on the anterior side with a
2.23 MHz linear probe (Figure I). We used a probe with a
0.73 inches diameter that is the most appropriate in terms
of positioning, location, orientation, and contact pressure.
Interobserver reproducibility of fat thickness measurements
with the ultrasound technique was good, and intraclass
correlation obtained by two examiners in the same subjects
was greater than 0.98.
Anthropometric measurements, weight, height, and,
umbilical waist circumference, were recorded by the same
operator using standard anthropometric techniques [20].
.. Statistics. Fat mass estimation by ultrasound uses a linear
regression equation developed from DXA reference value.
Tis regressionequationincludes the anthropometric charac-
teristics and subcutaneous fat thickness at the midthigh and
the back of the umbilical level. Statistical analysis was carried
out to compare the fat mass obtained by the ultrasound
technique and by DXAabsorptiometry using Student’s paired
?-test. Te accuracy of the fat mass estimations calculated
through the preceding models was evaluated from the deter-
mination coemcient ?
2
between true and estimated values
of fat mass and the standard error of the estimate (SEE)
as described by Lohman [2I]. Agreement between body
composition estimates was examined by calculating the 93%
limits of agreement as described by Bland and Altman [22].
Additionally, potential bias between BF estimates by DXA
and the US technique was obtained using residual plots. For
all analysis ? < 0.05 was considered signifcant. Statistical
tests were performed using Statistica sofware (version 6;
StatSof, Tulsa, Okla., USA).
3. Results
Te descriptive characteristics of the anthropometric mea-
surements, fat thicknesses measured by ultrasound and the
total fat mass (BF in kg) and BF% by DXA of sedentary
women and men separately are given in Table I.
Te regression equations used to estimate total fat mass
according to Hologic Discovery A are as follows.
Anutomlcul sltes used by Plneuu und colleugues
ln severul studles
How Ultrasound Works
• Cµerutes ln frequency runge: >z MEz
• Plezoelectrlc crystuls ln trunsducer of scun heud µroduce
µulses of ultrusound whlch ls then trunsmltted through
the skln.
• vhen beum comes ln contuct wlth u tlssue e.g.
subcutuneous fut or skeletul muscle lt ls µurtlully
refected buck to trunsducer us un echo. 1he umount of
sound thut ls refected buck deµends on lmµedunce.
• Alr hus c lmµedunce
• lut lmµedunce: c.¡¸8 gjcm
¡
js
¡

• Muscle lmµedunce: c.¡,c gjcm
¡
js
¡
• 8one lmµedunce: c.,8 gjcm
¡
js
¡

• 1he strength of euch wuve und µosltlon ls refected und
forms u dot. Lots come together und form un lmuge.
– Strong refectlons uµµeur
– veuk refectlons uµµeur 7'4,
– No echo: C#"+1
Limitations of Ultrasound
• Used slnce the mld¡ç6cs but $) 5&"$3"'3(L43 =')&)+)#
34M$43
• Lqulµment used vurles study to study:
– Mujorlty of studles revlewed used ¸ MEz, 8mode trunsducer
– 8odyMetrlx 8×zccc (Pro) mude by lnteluMetrlx: uses z.¸ MEz
– Plneuu et. ul used Amode trunsducer wlth z.z¸ MEz
• Muscle ls comµresslble when reluxed, dlmenslons of muscle chunge
wlth contructlonjreluxutlon
– 1oomey et ul (zc¡¡): Eulf mux force " SA1 thlckness by ¸6° Q trlceµ. lront thlgh " by
z¸°
• Muscle muss muy chunge wlth hydrutlon stutus - ,¸° E
z
c
• lut luyers ure not conslstent und cun vury whlle scunnlng
– 8ody Metrlx softwure uveruges for the user
– Ctherwlse, electronlc cullµers used, user must uveruge thlckness
• Cµtlmul meusurements tuken when trunsducer ls µurullel or trunsverse
to muscle
BF Estimate by Ultrasound
• lrom Plneuu et ul (zcc,)

• 94-,)4*
– 8l <µ¡ kg
• 8l (kg) = ¡.¸ 8Ml ÷ c.cµ Uthl ÷ c.zz Uthr ÷ c.c¡ Uµer - c.µ8
Mldthlgh thl ÷ c.µz Mldthlgh thr ÷ c.c¸6 Mldthlgh µer -
¡8.¸8
– 8l >µ¡ kg
• 8l (kg) = c.,6 8Ml - c.¸ Uthl ÷ c.ç Uthr ÷ c.µz Uµer = ¡.6z
Mld thlgh thl ÷ ¡.6z Mldthlgh thr ÷ ¡.¡z Mldthlgh µer - ç¸

8l = body fut estlmute, Uthl = umblllcul thlckness (left slde) ln mm, Uthr =
umblllcul thlckness (rlght slde) ln mm, Uµer = umblllcul µerlmeter ln cm, Mld
thlgh thl = mldthlgh thlckness (left slde) ln mm, Mldthlgh thr = mldthlgh
thlckness (rlght slde) ln mm, Mldthlgh µer = mld thlgh µerlmeter ln cm
BF Estimate by Ultrasound
• :,)4*
– ;9 <== >0
• ;9 ?>0@ A B.C= ;:D E B.BF G#$) E B.BH G#$' E B.=I GJ4' E B.B=C
:"KL#$"0$ #$) M B.BBN :"KL#$"0$ #$' M B.BOP :"KL#$"0$ J4' M
FH.O
– ;9 Q== >0
• ;9 ?>0@ A B.CN ;:D M B.CR G#$) E B.RC G#$' E B.FN GJ4' E B.=
:"KL#$"0$ #$) M B.RF :"KL#$"0$ #$' E B.BIR :"KL#$"0$ J4' M
CR.O

8l = body fut estlmute, Uthl = umblllcul thlckness (left slde) ln mm, Uthr =
umblllcul thlckness (rlght slde) ln mm, Uµer = umblllcul µerlmeter ln cm, Mld
thlgh thl = mldthlgh thlckness (left slde) ln mm, Mldthlgh thr = mldthlgh
thlckness (rlght slde) ln mm, Mldthlgh µer = mld thlgh µerlmeter ln cm

BodyMetrix Ultrasound Device
• Munufuctured by
lnteluMetrlx
– Meusures °8l, subcutuneous
und vlscerul fut lndeµendently
- uses u µroµrletury equutlon
– Llsted on Amuzon for sµç¸ for
ut home use
– "Pro' verslon uvulluble on
lnteluMetrlx webslte for
s¡ç8¸.cc lncludlng softwure
– Eus softwure whlch trucks
chunges ln body comµosltlon
over tlme (cun be udjusted
deµendlng on ftness level of
user)
CT vs. US
• 8uzzocchl et ul sought to determlne the correlutlon between US und C1
• All meusures hud meun P vulue <.ccc¡
• Mesenterlc fut thlckness wus only non slgnlfcunt meusure (P=.¡8)
• ln terms of reµroduclblllty - obese µoµulutlon (Crouµ 8, meun 8Ml of
¸,.¸) dlmcult versus Crouµ A (meun 8Ml of z,.µ), mesenterlc fut ulso hud
weuker reµroduclblllty (S A .II #& .BBFN@
be stressed that MAR-A and MAR-B again showed better
coefficients than WFI (r = 0.98–0.95 vs 0.90) in obese
patients.
DISCUSSION
The World Health Organization defines obesity as a patho-
logic condition, characterized by an excess of fat tissue in
the human body, which leads to a significant increase of
morbidity and mortality (46). The prevalence of obesity is
rising, with an alarming growth rate in all developed and
developing countries around the world; moreover, the
emerging or emerged problem of obesity during childhood
must be taken into consideration for future generations.
However, obesity is a heterogeneous and complex entity, in
particular with regard to its etiology and metabolic
complications.
In 1947, Vague (47) was the first to foresee that fat distribu-
tion could be more important than the total excess of fat. This
suggestion was later supported by several studies that stressed
the importance of the ‘‘central’’ distribution of fat mass in
the development of higher risk profile for cardiovascular
and metabolic diseases (48–50); the central phenotype is also
called ‘‘abdominal’’ or ‘‘android’’ because of the higher
prevalence in the male gender versus a peripheral
distribution (ie, ‘‘gynoid’’), which is more typical in women
(50). Moreover, in recent years, the development of tech-
niques such as CT imaging and MRI has allowed discrimina-
tion, with regard to the abdominal fat, between SATand VAT;
research has demonstrated that cardiometabolic fat–associated
risk is more linked to VAT than to SAT (8–12). The two fat
depots are distinct in their endocrine and paracrine
secretion profiles of hormones and cytokines; these
differences have an impact on glucose homeostasis. In
a study by Taksali et al (12), an adverse distribution of VAT
and SAT in adolescents was found to be related to a fivefold
greater risk for metabolic syndrome. Imaging techniques
have therefore enhanced measurement capabilities as well as
our understanding of the importance of VAT and have
shifted the attention of clinicians to the possibility not only
to detect the total fat mass of a patient (or the gynoid-
android distribution) but to define the distribution of fat accu-
mulation and its variation during, for instance, therapy or
follow-up. CTimaging and MRI are unanimously considered
the gold standards for the assessment of total and regional
adiposity, but for evident reasons (availability, costs, radiation
exposure for CT), their use is definitely limited to research;
nevertheless their support could be considered when a patient
undergoes these ‘‘heavy’’ imaging techniques for different and
‘‘clinically relevant’’ reasons to provide fast measurements or
indexes of adiposity aside fromthe main problemof the exam-
ination. Hence, interest by clinicians and researchers is
increasing to find suitable methods for the analysis of BC
that may represent an alternative for the assessment of regional
adiposity in clinical use.
BMI has been used by the World Health Organization as
the standard for recording obesity statistics since the early
1980s and represents the most used index of relative body
weight. Individuals whose BMIs are #18.5 kg/m
2
are consid-
ered underweight, whereas those whose BMIs range from
18.5 to 24.9 kg/m
2
are classified as having normal or accept-
able weight. Those whose BMIs range from 25.0 to 29.9 kg/
m
2
are commonly referred to as overweight, and obesity is said
to be present when BMI is >30.0 kg/m
2
. Three grades of
obesity are defined on the basis of BMI: grade 1 (30.0–34.9
kg/m
2
), grade 2 (35.0–39.9 kg/m
2
), and grade 3 ($40.0
Figure 3. (a) Accuracy: correlation between
radiologist 1 (RAD 1) ultrasound (US) measure-
ments and computed tomographic (CT)
measurements for intra-abdominal fat thick-
ness, group A. (b) Reproducibility: correlation
between radiologist 2 (RAD 2) and radiologist 3
(RAD3) US measurements for maximumsubcu-
taneous fat thickness B, group B. (c) Reproduc-
ibility: intraclass correlation among all
radiologists for intra-abdominal fat thickness,
group B. (d) Repeatability: correlation between
two independent measurements of maximum
preperitoneal fat thickness by RAD 1, group B.
BAZZOCCHI ET AL Academic Radiology, Vol 18, No 9, September 2011
1138
be stressed that MAR-A and MAR-B again showed better
coefficients than WFI (r = 0.98–0.95 vs 0.90) in obese
patients.
DISCUSSION
The World Health Organization defines obesity as a patho-
logic condition, characterized by an excess of fat tissue in
the human body, which leads to a significant increase of
morbidity and mortality (46). The prevalence of obesity is
rising, with an alarming growth rate in all developed and
developing countries around the world; moreover, the
emerging or emerged problem of obesity during childhood
must be taken into consideration for future generations.
However, obesity is a heterogeneous and complex entity, in
particular with regard to its etiology and metabolic
complications.
In 1947, Vague (47) was the first to foresee that fat distribu-
tion could be more important than the total excess of fat. This
suggestion was later supported by several studies that stressed
the importance of the ‘‘central’’ distribution of fat mass in
the development of higher risk profile for cardiovascular
and metabolic diseases (48–50); the central phenotype is also
called ‘‘abdominal’’ or ‘‘android’’ because of the higher
prevalence in the male gender versus a peripheral
distribution (ie, ‘‘gynoid’’), which is more typical in women
(50). Moreover, in recent years, the development of tech-
niques such as CT imaging and MRI has allowed discrimina-
tion, with regard to the abdominal fat, between SATand VAT;
research has demonstrated that cardiometabolic fat–associated
risk is more linked to VAT than to SAT (8–12). The two fat
depots are distinct in their endocrine and paracrine
secretion profiles of hormones and cytokines; these
differences have an impact on glucose homeostasis. In
a study by Taksali et al (12), an adverse distribution of VAT
and SAT in adolescents was found to be related to a fivefold
greater risk for metabolic syndrome. Imaging techniques
have therefore enhanced measurement capabilities as well as
our understanding of the importance of VAT and have
shifted the attention of clinicians to the possibility not only
to detect the total fat mass of a patient (or the gynoid-
android distribution) but to define the distribution of fat accu-
mulation and its variation during, for instance, therapy or
follow-up. CTimaging and MRI are unanimously considered
the gold standards for the assessment of total and regional
adiposity, but for evident reasons (availability, costs, radiation
exposure for CT), their use is definitely limited to research;
nevertheless their support could be considered when a patient
undergoes these ‘‘heavy’’ imaging techniques for different and
‘‘clinically relevant’’ reasons to provide fast measurements or
indexes of adiposity aside fromthe main problemof the exam-
ination. Hence, interest by clinicians and researchers is
increasing to find suitable methods for the analysis of BC
that may represent an alternative for the assessment of regional
adiposity in clinical use.
BMI has been used by the World Health Organization as
the standard for recording obesity statistics since the early
1980s and represents the most used index of relative body
weight. Individuals whose BMIs are #18.5 kg/m
2
are consid-
ered underweight, whereas those whose BMIs range from
18.5 to 24.9 kg/m
2
are classified as having normal or accept-
able weight. Those whose BMIs range from 25.0 to 29.9 kg/
m
2
are commonly referred to as overweight, and obesity is said
to be present when BMI is >30.0 kg/m
2
. Three grades of
obesity are defined on the basis of BMI: grade 1 (30.0–34.9
kg/m
2
), grade 2 (35.0–39.9 kg/m
2
), and grade 3 ($40.0
Figure 3. (a) Accuracy: correlation between
radiologist 1 (RAD 1) ultrasound (US) measure-
ments and computed tomographic (CT)
measurements for intra-abdominal fat thick-
ness, group A. (b) Reproducibility: correlation
between radiologist 2 (RAD 2) and radiologist 3
(RAD3) US measurements for maximumsubcu-
taneous fat thickness B, group B. (c) Reproduc-
ibility: intraclass correlation among all
radiologists for intra-abdominal fat thickness,
group B. (d) Repeatability: correlation between
two independent measurements of maximum
preperitoneal fat thickness by RAD 1, group B.
BAZZOCCHI ET AL Academic Radiology, Vol 18, No 9, September 2011
1138
Acud Rudlol zc¡¡, ¡8:¡¡¸¸¡¡µ¸
Thickness Measurement CT vs. US
same ultrasound measures were also derived on CT axial
and multiplanar sagittal reconstructed images (acquisition
1.25 mm, acquired before any contrast administration) by an
independent radiologist.
Abdominal Adiposity Measures
Thickness measurements. Maximum preperitoneal fat thick-
ness was assessed using a linear probe, longitudinal scan on
the xiphoumbilical line, measured just below the xiphoid
process in the epigastric region as the major distance between
the anterior surface of the peritoneum covering the liver (left
lobe) to the posterior surface of linea alba (Figs 1a and 2a).
Minimum subcutaneous fat thickness was assessed using
a linear probe, longitudinal scan on the xiphoumbilical line,
in the same anatomic place of maximum preperitoneal fat
thickness (just below the xiphoid process in the epigastric
region) determined as the distance between the anterior
surface of linea alba and the fat-skin barrier (Figs 1a and 2a).
Maximum subcutaneous fat thickness was assessed using
a linear probe, longitudinal scan, evaluated with two measure-
ments: maximum subcutaneous fat thickness A was measured
2 cm above the umbilicus (Figs 1b and 2b) and maximum
subcutaneous fat thickness B 2 cm below the umbilicus
(Fig 2c), both measuring the distance between the linea alba
and the fat-skin barrier. Maximum subcutaneous fat thickness
B has not been described in previous literature.
Intra-abdominal fat thickness was assessed using a convex
probe, longitudinal scan on the xiphoumbilical line, 2 cm
above the umbilicus, measured as the distance between the
anterior wall of the aorta and the posterior surface of the
rectus abdominis muscle (Figs 1c and 2b).
Mesenteric fat thickness was assessed using a convex probe,
transversal scan in the periumbilical area, measured as the
maximum distance between three contiguous mesenteric
leaves. The mesenteric leaves appear to be elongated structures
with highly reflecting peritoneal surfaces; they have high-level
echoes in the periphery, and small vascular structures (1–2 mm
in diameter) may be seen within it. Measurements of three
different distances between mesenteric leaves were obtained
on the basis of anatomic landmarks as described by other
investigators (15,30), and the mean value was used for analysis.
Adiposity indexes. The following adiposity indexes were
derived from ultrasound thickness measurements: wall fat
index (WFI), ratio between minimumsubcutaneous fat thick-
ness and maximum preperitoneal fat thickness, maximum
abdominal ratio A (MAR-A) as the ratio between maximum
subcutaneous fat thickness A and intra-abdominal fat thick-
ness, and maximum abdominal ratio B (MAR-B) as the ratio
between maximum subcutaneous fat thickness B and intra-
abdominal fat thickness, both never considered in previous
studies.
Statistical Analysis
Statistical analysis was performed using MedCalc version
11.4.2 (MedCalc Software, Mariakerke, Belgium).
Figure 1. Examples of different computed tomographic (CT) and ultrasound measurements. (a) Minimum subcutaneous and maximum pre-
peritoneal fat thickness obtained by CT imaging (a1 and a2, respectively) and ultrasound (a3 and a4, respectively). (b) Maximumsubcutaneous
fat thickness A obtained by CT imaging (b1) and ultrasound (b2). (c) Intra-abdominal fat thickness obtained by CT imaging (c1) and ultrasound
(c2).
Academic Radiology, Vol 18, No 9, September 2011 ULTRASONOGRAPHY IN ABDOMINAL ADIPOSITY
1135
Acud Rudlol zc¡¡, ¡8:¡¡¸¸¡¡µ¸
Conclusions
• ve know body comµosltlon ls lmµortunt -
therefore we must do our best to meusure lt
• Luch method hus lts own dlsudvuntuges und
udvuntuges
– Mujor dlsudvuntuge ln uslng C1jUS: no consensus ln
llteruture on uslng these tools to determlne ureu,
volume or µercentuges
• ln C1 - Sllceomutlc seemed to be conslstent ln most recent
studles
• US - equlµment used vurled study to study, brlef overvlew
of studles done for uthletes used 8odyMetrlcs
– Softwure seems to mlnlmlze user error und lncreuse
reµroduclblllty rutes
References
Antoun, S., Lunoy, L., AlblgesSuuvln, L., & Lscudler, 8. (zc¡µ) Cllnlcul lmµllcutlons of
body comµosltlon ussessment by comµuted tomogruµhy ln metustutlc renul cell
curclnomu. TUJ4'# V43"4! &+ WX#"1,X14' Y$4',J5Z FP(¸), z,çz88.

8urucos, v.L. Meusurement of leun body muss uslng C1 scuns. Retrleved from:
httµ:jjstutlc.ubbottnutrltlon.comjcmsµrodjunhl.orgjlmgj¡ç¡¡cth°zcAN°zcConf
°zc8urucos°zcllnul¡¡cth°zcAbbott°zcNutrltlon°zcReseurch°zcConference.µdf.

8urucos, v.L., Relmun, R., Mourtzukls, M., Cloulbusunls, l., & Antoun, S. (zc¡c). 8ody
comµosltlon ln µutlents wlth nonsmull cell lung cuncer: u contemµorury vlew of
cuncer cuchexlu wlth the use of comµuted tomogruµhy lmuge unulysls. W-4'"1,X
[&('X,) &+ \)"X"1,) ](#'"#"&XZ OFZ ¡¡¸¸S,S.

8uzzocchl, A., lllonzl, C., Pontl, l., Sussl, C,. Sullzzonl, L., 8uttlstu, C., & Cunlnl, R.
(zc¡¡). Accurucy, reµroduclblllty und reµeutlublllty of ultrusonogruµhy ln the
ussessment of udomlnul udlµoslty. W1,K4-"1 V,K"&)&05Z FNZ ¡¡¸¸¡¡µ¸.

8onekumµ, S., Chosh, P., Cruwford, S., Solgu, S.l., Eorsku, A., 8runcutl, l.L., Llehl,
A.M . Clurk, '.M. (zcc8). Quuntltlutlve comµurlson und evuluutlon of softwure
µuckuges for ussessment of ubdomlnul udlµose tlssue dlstrlbutlon by mugnetlc
resonunce lmuglng. DX#4'X,#"&X,) [&('X,) &+ 674*"#5Z C=, ¡cc¡¡¡.
References
lunelll, M.1., & luczmurskl, R.'. (¡ç8µ). Ultrusound us un uµµrouch to
ussesslng body comµosltlon. W-4'"1,X [&('X,) &+ \)"X"1,) ](#'"#"&XZ COZ
,c¸,cç.

'ohnson, l.L., Nuccuruto, l.A., Corder, M.A., & Reµovlch, v.L. (zc¡z).
vulldutlon of three body comµosltlon technlques wlth u comµurlson of
ultrusound ubdomlnul fut deµths ugulnst un octoµolur bloelectrlcul
lmµedunce devlce. DX#4'X,#"&X,) [&('X,) &+ TU4'1"*4 ^1"4X14Z R(¸), zc¸z¡¸.

lvlst, E., Chowdhury, 8., Crungurd, U., 1ylen, U., & Sjostorm, L. (¡ç88).
1otul und vlsercul udlµosetlssue volumes derlved from meusurements wlth
comµuted tomogruµhy ln udult men und women: µredlctlve equutlons.
W-4'"1,X [&('X,) &+ \)"X"1,) ](#'"#"&XZ PNZ ¡¸¸¡6¡.

luwukuml, Y., Aklmu, E., lubo, l., Muruoku, Y., Euseguwu, E., louzukl, M.,
lmul, M . lukunugu, 1. (zcc¡). Chunges ln muscle slze, urchltecture und
neurul uctlvutlon ufter zc duys of bed rest wlth und wlthout reslstunce
exerclse. [&('X,) &+ WJJ)"4K S$5*"&)&05Z NP(¡z), ,¡z.
References
Muyuns, L., Curtwrlght, M., & vulker, l.C. (zc¡z). Neuromusculur
ultrusonogruµhy: quuntlfylng muscle und nerve meusurements. S$5*"1,)
:4K"1"X4 _ V4$,7")"#,#"&X \)"X"1* &+ W-4'"1,Z =CZ ¡¸¸¡µ8.

Mlcklesfeld, L.l., Coedecke, '.E., Punyunltyu, M., vllson, l.L., & lelly, 1.L.
(zc¡z). Luulenergy xruy µerforms us well us cllnlcul comµuted tomogruµhy for
the meusurement of vlscerul fut. 674*"#5Z =BZ ¡¡cç¡¡¡µ. dol:¡c.¡c¸8joby.zc¡¡.¸6,

'ensen, M.L., lunuley, '.A., Reed, '.L., & Sheedy, P.l. (¡çç¸). Meusurement of
ubdomlnul und vlscerul fut wlth comµuted tomogruµhy und duulenergy xruy
ubsorµtlometry. W-4'"1,X [&('X,) &+ \)"X"1,) ](#'"#"&XZ IFZ z,µ8.

Plneuu, '.C., Lulys, L., Pellegrlnl, M., & 8uttlstlnl, N.C. (zc¡¸). 8ody fut muss
ussessment: u comµurlson between un ultrusoundbused devlce und u
dlscovery A model of L×A. [&('X,) &+ 674*"#5Z dol:¡c.¡¡¸¸jzc¡¸jµ6z¸çµ

Plneuu, '.C., CulhurdCostu, A.M., & 8ocquet, M. (zcc,). vulldutlon of
ultrusound technlques uµµlled to body fut meusurement. ](#'"#"&X _
:4#,7&)"*-Z RFZ µz¡µz,.
References
Reld, C.L., Cumµbell, l.1., Llttle, R.A. (zcc¸). Muscle wustlng und
energy bulunce ln crltlcul lllness. \)"X"1,) ](#'"#"&XZ =C(z), z,¸z8c.

Rlberlolllho, l.l., lurlu, A.N., lohlmunn, C., Ajzen, S., Rlberlo, A.8.,
2unellu, M.1., & lerrelru, S.R.C. (zcc¡). Ultrusonogruµhy for the
evululutlon of vlscerul fut und curdlovusculur rlsk. /5J4'#4X*"&XZ CNZ
,¡¸,¡,

Smlth, S.R., Lovejoy, '.C., Creenwuy, l., Ryun, L., de'onge, L., de lu
8retonne, ' . 8ruy, C.A. (zcc¡). Contrlbutlons of totul body fut,
ubdomlnul subcutuneous udlµose tlssue comµurtments und vlscerul
udlµose tlssue to the metubollc comµllcutlons of obeslty. :4#,7&)"*-Z
RB(µ), µz¸µ¸¸.

1oomey, C., McCreesh, l., Leuhy, S., & Lukemun, P. (zc¡¡). 1echnlcul
conslderutlons for uccurute meusurement of subcutuneous udlµose
tlssue thlckness uslng 8mode ultrusound. G)#',*&(XKZ FOZ ç¡ç6.

vugner, L.R. (zc¡¸). Ultrusound us u tool to ussess body fut. [&('X,) &+
674*"#5Z dol:¡c.¡¡¸¸jzc¡¸jz8c,¡¸