- Email: [email protected]
The Role of Ultrasonography in the Evaluation of Abdominal Fat
The Role of Ultrasonography in the Evaluation of Abdominal Fat: Analysis of Technical and Methodological Issues Alberto Bazzocchi, MD, PhD, Giacomo Filonzi, MD, Federico Ponti, MD, Michele Amadori, MD, Claudia Sassi, MD, Eugenio Salizzoni, MD, Ugo Albisinni, MD, Giuseppe Battista, MD Rationale and Objectives: Ultrasonography (US) is becoming popular for the assessment of adiposity, but no one has studied this tool in the light of its potential limitations. Our purpose was to investigate the impact of technical conditions on the evaluation of abdominal fat by US. Materials and Methods: Forty-five healthy males and 45 healthy females were consecutively enrolled in the study, randomly assigned to three groups equally distributed by sex, and examined accordingly to three technical points: fasting state (before and after meal [A]), breathing (expiration and inspiration [B]), and US equipment from different generations: 2003 and 1998 (C). Two blinded radiologists performed US in the these opposite conditions, acquiring five parameters representative of subcutaneous and visceral adiposity in two times. Student’s t-test and Lin’s correlation coefficient were used for statistical analysis to assess differences in the measures as well as in inter- and intra-observer agreements. Results: The maximum and the only statistically significant changes were observed for intra-abdominal fat thickness regarding fasting state and breathing (D% = 24.1 21.3 and D% = 9.2 20.4, respectively; P < .0001). Reproducibility and repeatability, especially for visceral fat, were proved more stable in the following conditions: fasting state, expiration, and newer machine (2003). Conclusion: This article provides essential information and ‘‘range of confidence’’ for variations that can be expected from using different conditions in the measurement of abdominal adiposity by US to be carefully addressed as well as considered by US users and by researchers involving this technique in the field of body composition. Key Words: Body composition; adiposity; reproducibility of results; methods; ultrasonography. ªAUR, 2013
A
dipose tissue is one of the most enigmatic components of the body. Evaluation and quantification of adipose tissue are fundamental in the field of body composition as well as body composition is crucial for understanding human and animal metabolism and its alterations (1). Imaging has been essential in bringing body composition analysis to a clinical ground. In clinical practice, one of the most used techniques to assess the organ-tissue level of body composition is ultrasonography (US). This technique has demonstrated satisfying results in terms of accuracy and reproducibility (2–11). Moreover, US is ‘‘friendly’’ and oriented to one of the key points in the assessment of fat tissue: its distribution between subcutaneous and visceral compartments. The importance of this point has already been proven (12–17). Risk factors
Acad Radiol 2013; 20:1278–1285 From the Department of Specialized, Diagnostic, and Experimental Medicine, University of Bologna, Sant’Orsola - Malpighi Hospital, Via G. Massarenti 9, 40138 Bologna, Italy (A.B., G.F., F.P., M.A., C.S., E.S., G.B.); and Diagnostic and Interventional Radiology, ‘‘Rizzoli’’ Orthopaedic Institute, Bologna, Italy (A.B., U.A.). Received April 14, 2013; accepted July 24, 2013. Address correspondence to: A.B. e-mail: [email protected] ªAUR, 2013 http://dx.doi.org/10.1016/j.acra.2013.07.009
1278
for complications associated with obesity and the wider field of ‘‘metabolic diseases’’ start from fat, whether it is visceral or not. Recently, some studies underlined that the ratio between subcutaneous and visceral fat could be an important index of risk for impaired glucose tolerance, hypertriglyceridemia, hypercholesterolemia, dyslipidemia, fatty liver, and liver dysfunction (18,19), and adolescents with a prevalent visceral distribution were found to have a five times higher risk of developing metabolic syndrome than those who had a subcutaneous distribution (20). Although few works previously proposed (2–11), validated, and ‘‘consecrated’’ US for fat evaluation of different body sites, mainly of the abdomen, no one considered investigating US for important technical and methodological concerns. How might abdominal distension influence US assessment of adiposity? How might thoracic and abdominal movements during respiration affect the measurements? Does the use of older machines instead of new ones affect results? None of these points have yet been considered or defined. Because US, in this field, is advocated not for the qualitative evaluation of a tissue or an organ, but for the quantification of a tissue or an organ, all of these issues may be even more important. An error in measurement may represent a huge problem when comparing different patients, or patients at
Academic Radiology, Vol 20, No 10, October 2013
different time points in their clinical follow-up. It might be also interesting to know when and how these errors may occur and the cause of the variations in the measurements. Furthermore, all previous studies investigating US in this field showed inhomogeneous criteria in the assessment of abdominal adiposity and many variations in methods (Table 1). The aim of our study was to investigate the impact of abdominal distension, patients’ breathing, and US equipment in the evaluation of subcutaneous and visceral abdominal fat by US. MATERIALS AND METHODS Study Design and Population
Ninety healthy volunteers (45 males and 45 females) were consecutively enrolled in the study. Exclusion criteria were: history of pathologies involving the trunk (thorax and abdomen) and potentially affecting its anatomy, including past surgical treatments or implanted devices. The patients were randomly assigned to three groups and equally distributed by sex (15 males and 15 females for each group). Subjects in group A (32.9 5.4 years old; body mass index [BMI] 24.1 2.7 kg/m2, range 21.2–28.9) were submitted to US examination (Esaote Technos MPX-2003; Esaote, Genoa, Italy) before (in a fasting state of at least 12 hours) and 45 minutes after a standardized meal (lunch), with measurements performed during a mild expiration. Subjects of group B (29.8 3.8 years old; BMI 23.6 2.9 kg/m2, range 18.4–28.2) underwent US assessment in mild expiration and 30 minutes later in mild inspiration, both times in a fasting state (at least 12 hours) and with the same US machine as for group A. Group C (33.5 4.1 years old; BMI 25.3 2.6 kg/m2, range 20.4–29.7) was evaluated in a fasting state (at least 12 hours) and during a mild expiration by using US equipments of different generations: equipment 1 (Esaote Technos MPX; year 2003; tissue harmonic imaging, yes; compounding, no; maximum scan depth, 28.5 cm; gray levels, 256) and equipment 2 (Hitachi EUB 525 eidos; Hitachi Denshi, Tokyo, Japan; year, 1998; tissue harmonic imaging, no; compounding, no; maximum scan depth, 24 cm; gray levels, 256). Thus, patients of all groups were evaluated in two different conditions: (a) fasting state, (b) breathing, and (c) equipment. Moreover these conditions were all tested by two skilled radiologists twice to evaluate inter- and intra-observer variability of the measurements associated with the specified conditions. The ‘‘state-of-the-art’’ evaluation (patients in group C, equipment 1) was also repeated twice, with the lowest and the highest hand (probe) pressure on the abdomen to verify the potential influence of this element. To blind the study, all measurements were independently collected by a third physician, and the readers were blinded to each others in the US scans. The study was carried out respecting the principles of the Declaration of Helsinki. The study protocol was approved by the local institutional
ADIPOSITY: ISSUES FOR ULTRASOUND IMAGING
review board and all participants signed an informed consent form. Ultrasound Methods
All measurements were performed with subjects in a supine position with arms at sides. Five parameters, representative of subcutaneous and visceral fat, were considered in the three groups: 1) minimum subcutaneous fat thickness (MinSFT); maximum subcutaneous fat thickness, upper (MaxSFTupper); and maximum subcutaneous fat thickness, lower (MaxSFTlower); 2) maximum preperitoneal fat thickness (MaxPFT); and 3) intra-abdominal fat thickness (IFT). MaxPFT measurement was performed just below the xiphoid process in the epigastric region, with a longitudinal scan on the xiphoumbilical line, as the major distance between the anterior surface of the peritoneum covering the liver (left lobe) and the posterior surface of linea alba. MinSFT was measured in the same anatomic region as the maximum preperitoneal fat. MaxSFT was assessed at two different sites on the linea alba, 2 cm above and 2 cm underneath the umbilicus (MaxSFTupper and MaxSFTlower, respectively), with a longitudinal scan. MinSFT, MaxSFTupper, and MaxSFTlower were defined as the distance between the anterior surface of linea alba and the fat-skin barrier. IFT was measured as the distance between the anterior wall of the aorta and the posterior surface of linea alba, 2 cm above the umbilicus, with a transverse scan (Fig 1). MaxPFT, MinSFT, MaxSFTupper, and MaxSFTlower were measured using a linear probe (7.5 MHz) kept perpendicular to the skin and with hand pressure on the abdomen as light as possible in order not to compress the fat layers (with the exception of the circumstance in group C); IFT was assessed using a convex probe (3.5 MHz) with the same modalities. All of these parameters were investigated by our team in 2011 in terms of accuracy, inter-observer, and intraobserver agreements in people with normal and high BMI (9). Statistical Analysis
The normal distribution of our sample population was tested by skewness and excess kurtosis test. Normal ranges were considered for values between 2 and +2. Results are reported as mean and standard deviation (SD), and integrated with Delta (D) and Delta percentage (D%). Student’s t-test was used to analyze the statistical impact of abdominal distension, respiration, and equipment on US abdominal adiposity measurements; P was considered significant for values less than .05. Lin’s correlation coefficient (r) was used to assess intra- and inter-operator agreements in all groups for all conditions and parameters. A poor correlation was assumed for r values less than 0.70; moderate correlation with r between 0.70 and 0.79; good correlation with r between 0.80 and 0.89; very good correlation between 0.90 and 0.94; and excellent correlation with r of 0.95 or more (21). A ‘‘significant’’ difference in reproducibility and 1279
BAZZOCCHI ET AL
Academic Radiology, Vol 20, No 10, October 2013
TABLE 1. Methods and Technical Issues of the Principal Articles Measuring Visceral Fat Thickness with US Between 2001 and 2011
Author Publication
Fasting Condition
Breath Condition
Emmons RR, Ultrasound Med Biol (2011)
Not declared
Upon exhalation
Oh J, Ultrasound Med Biol (2011)
After fasting at least for 4 hours
Immediately After expiration
Anatomical Landmarks of Visceral Fat From the linea alba to the anterior aorta
From the anterior wall of the aorta to the internal face of the rectus abdominis muscle De Lucia Rolfe E, Refrain from eating At the end of a quiet From the peritoneal boundary Obesity (2010) 10 hours before expiration to the corpus of the lumbar vertebra Gradmark AM, Br J Fasting state Hold breath during From the inside of the bowel Nutr (2010) mid-inspiration wall to the spine Bartha JL, Obesity Fasting state Immediately at the From the anterior wall of the (2007) end of expiration aorta to the internal layer of the rectoabdominal muscle Koda M, Abdom Fasting overnight Not declared From the internal face of the Imaging (2007) rectus abdominis muscle (linea alba) to the anterior wall of the vertebra at the level of the umbilicus Not declared Breath-hold From the internal face of the Guldiken S, Int J Clin Pract (2006) abdominal muscle to the anterior wall of the aorta Kim SK, Am J Clin Not declared Immediately after From the anterior wall of the Nutr (2004) respiration aorta to the internal face of the rectoabdominal muscle Ribeiro-Filho FF, Fasting state Not declared From the internal face of Obes Res (2003) rectoabdominal muscle to the anterior wall of the aorta Leite CC, Metabolism Not declared Not declared From the internal face of the (2002) abdominal muscles to the posterior wall of the aorta Stolk RP, Int J Obes Relat Not declared At the end of a quiet From the posterior edge of the Metab Disord (2001) expiration abdominal muscles to the lumbar spine or psoas muscles Sabir N, Eur J Ultrasound Not declared Breath-hold From the internal face of the (2001) abdominal muscle to the anterior wall of the aorta.
repeatability between the measurements was defined for results (r) belonging to different classes of agreements (poor, moderate, etc.). Statistical analysis was performed using MedCalc version 11.4.2 (MedCalc Software, Mariakerke, Belgium).
US Equipment GE Logiq Book XP (GE Healthcare, Buckinghamshire, UK) Prosound a10 (Aloka, Tokyo, Japan) Logic Book XP ultrasound (GE Healthcare, Bedford, UK) Acuson Sequoia, (Siemens, Mountain Woods, California) Esaote Technos MPX (Esaote S.p.A., Genova, Italy) Aloka SSD-2000 (Aloka, Tokyo, Japan)
Sonoline Elegra, (Siemens Medical Solutions, Inc., Issaquah, Washington) SA 9900 (Medison, Seoul, Korea) Not declared
Toshiba Sonolayer SSA-250 A (Otawara Shi, Tochigiken, Japan) ATL HDI 3000 system (Bothell, Washington)
GE logiq 200
% = 24.1 21.3, absolute value D% = 25.7 21.2, minimum D% = 13.3, maximum D% = 73.7, P < .0001) (Fig 2). Subcutaneous measures and MaxPFT were not affected by abdominal distension, at least not with a statistical significance (Fig 3).
US values and differences for technical conditions are presented in Table 2.
Group B. As for group A, IFT showed the most important variation (D 5.6 8.2 mm, D% = 9.2 20.4, absolute value D% = 18.75 11.91, minimum D% = 31.4, maximum D% = 42.6, P < .0001). On the other side, none of the other parameters showed significant variation between the two different conditions.
Group A. IFT showed important variations before and after the meal, from 47.4 17.3 mm to 56.9 15.2 mm (D
Group C. Notwithstanding the important technological gap between the two US machines, no significant difference
RESULTS Descriptive Statistics
1280
Academic Radiology, Vol 20, No 10, October 2013
Figure 1. Abdominal sagittal section showing anatomical landmarks and parameters of adiposity as evaluated by ultrasonography: minimum subcutaneous fat thickness (a); maximum preperitoneal fat thickness (b); maximum subcutaneous fat thickness, upper (c); intra-abdominal fat thickness (d); and maximum subcutaneous fat thickness, lower (e).
resulted in the measurements performed by the two machines. The level of tissue compression by the probe produced extremely relevant changes of values for all parameters (D% ranging between 20.6% and 47.7% for subcutaneous parameters and between 53.3% and 74.8% for visceral parameters).
Analysis of Reliability
All data are shown in Table 3 (reproducibility) and Table 4 (repeatability).
Reproducibility. All parameters showed moderate to excellent reproducibility in all conditions evaluated (r between 0.78 and 0.95). IFT, MaxPFT, and MinSFT were the only parameters affected by statistically relevant changes in reproducibility. IFT showed significant variations of interobserver agreement in all groups (r 0.92–0.85 in group A, 0.91–0.86 in group B, 0.95–0.87 in group C). MaxPFT and MinSFT were also critically unstable, but only for group B (r 0.83–0.78 and 0.92–0.86, respectively).
Repeatability. All measurements were reported with moderate to excellent values of repeatability, for both operators (r ranging between 0.77 and 0.99). However, IFTwas again affected by statistically considerable changes, although these were more dependent on single operators in groups A and B.
TABLE 2. Ultrasound Values, Delta (D), and Delta Percentage (D%) in Group A (Abdominal Distension), Group B (Respiration), and Group C (Equipment)
MinSFT MaxPFT MaxSFTupper MaxSFTlower IFT
Respiration
Equipment
Pre-m (mm) Mean SD min/max
Post-m (mm) D (mm) D% (%) Exp (mm) Mean SD Mean SD Mean SD Mean SD min/max min/max min/max min/max
Insp (mm) Mean SD min/max
D (mm) Mean SD min/max
D% (%) Mean SD min/max
E 1 (mm) E 2 (mm) D (mm) Mean SD Mean SD Mean SD min/max min/max min/max
7.6 3.4 3.7/14.6 14.3 4.2 5.3/22.4 18.8 7.2 6.9/31.4 19.7 9.1 5.6/40.8 47.4* 17.3 27.3/92.4
7.6 2.9 3.0/14.7 15.0 4.9 5.5/21.9 18.8 7.1 6.5/29.0 20.1 9.0 6.9/39.0 56.9* 15.2 35.9/95.6
5.8 2.4 2.7/10.3 13.6 3.9 7.9/21.2 16.2 6.2 6.2/26.8 17.6 7.6 6.5/33.1 43.6* 12.8 20.3/75.1
0.1 1.1 3.2/3.0 0.3 2.2 3.9/5.9 0.5 1.8 3.2/3.8 0.5 1.9 5.4/3.4 5.6 8.2 11.2/15.2
0.6 20.2 34.0/42.2 4.2 15.2 40.3/27.8 4.1 11.4 27.4/16.3 4.6 11.8 10.6/26.3 9.2 20.4 31.4/42.6
8.6 3.0 8.4 2.8 3.1/15.9 3.4/15.2 14.5 5.2 13.5 5.1 5.3/22.4 5.0/21.3 19.5 7.9 19.2 8.0 7.9/31.5 6.9/29.5 21.1 8.9 21.9 9.8 5.7/40.8 7.6/41.8 50.9 14.1 51.3 16.0 31.4/82.1 27.3/88.4
0.0 1.8 3.7/7.7 0.8 1.7 3.2/6.2 0.0 2.1 6.0/5.8 0.4 2.2 4.2/7.3 9.5 8.1 14.2/25.6
1.0 22.8 54.4/53.4 4.8 12.5 20.3/40.2 0.4 11.5 32.3/31.5 4.0 13.9 18.8/42.6 24.1 21.3 15.3/73.7
5.7 2.7 5.8/12.6 13.9 3.7 7.9/21.8 16.7 6.0 7.0/27.1 18.1 7.3 8.7/34.3 38.0* 9.9 21.9/59.1
0.1 0.9 2.2/2.0 1.0 2.4 10.5/1.1 0.2 2.5 8.7/3.5 0.8 2.9 7.7/7.3 0.4 4.1 8.6/9.6
D% (%) Mean SD min/max 3.6 11.4 30.5/13.4 10.2 20.4 68.2/5.6 2.2 16.5 65.9/16.6 6.0 15.6 37.0/34.1 0.1 10.2 27.3/26.0
1281
E 1, equipment 1; E 2, equipment 2; exp, expiration; IFT, intra-abdominal fat thickness; insp, inspiration; max, maximum; MaxPFT, maximum preperitoneal fat thickness; MaxSFTlower, maximum subcutaneous fat thickness 2 cm underneath the umbilicus; MaxSFTupper, maximum subcutaneous fat thickness 2 cm above the umbilicus; min, minimum; MinSFT, minimum subcutaneous fat thickness; pre-m, premeal; post-m, postmeal. *Student’s t-test: significant change with P < .005.
ADIPOSITY: ISSUES FOR ULTRASOUND IMAGING
Abdominal Distension
BAZZOCCHI ET AL
Academic Radiology, Vol 20, No 10, October 2013
Figure 2. Opposite conditions for intraabdominal fat thickness in three subjects, one for each group (a, b, and c); before (a1) and after (a2) food consumption; in expiration (b1) and inspiration (b2); and acquired by equipment 1 (c1) and equipment 2 (c2). Thicknesses are expressed in millimeters.
Figure 3. Opposite conditions in three subjects, one for each group (a, b, and c) for subcutaneous fat parameters and maximum preperitoneal fat thickness: minimum subcutaneous fat thickness and maximum preperitoneal fat thickness before (a1) and after (a2) food consumption; maximum subcutaneous fat thickness, upper, during expiration (b1) and during inspiration (b2); and maximum fat thickness, lower, with equipment 1 (c1) and equipment 2 (c2). Thicknesses are expressed in millimeters.
TABLE 3. Reproducibility of Sonographic Measurements in Group A (Abdominal Distension), Group B (Respiration), and Group C (Equipment) Abdominal Distension
MinSFT MaxPFT MaxSFTupper MaxSFTlower IFT
Respiration
Equipment
Premeal
Postmeal
Expiration
Inspiration
Equipment 1
Equipment 2
0.86 0.82 0.90 0.91 0.92*
0.83 0.85 0.94 0.93 0.85*
0.92* 0.83* 0.95 0.94 0.91*
0.86* 0.78* 0.95 0.91 0.86*
0.87 0.84 0.94 0.92 0.95y
0.81 0.81 0.93 0.94 0.87y
IFT, intra-abdominal fat thickness; MaxPFT, maximum preperitoneal fat thickness; MaxSFTupper, maximum subcutaneous fat thickness 2 cm above the umbilicus; MaxSFTlower, maximum subcutaneous fat thickness 2 cm underneath the umbilicus; MinSFT, minimum subcutaneous fat thickness. *One class of agreement gap between r. y Two classes of agreement gap between r.
MinSFT showed significant variations in all groups, whereas MaxPFT showed important variations in group A and group C. All other parameters suffered from few variations, occasional variations, or no variation at all. The compression of the abdomen did not statistically influence the level of reproducibility and repeatability. 1282
DISCUSSION Today, US is one of the most suitable tools for evaluating abdominal adiposity in clinical practice, especially in the differential assessment of subcutaneous and visceral adipose depots. Some of the most attractive features of US—such as
IFT, intra-abdominal fat thickness; MaxPFT, maximum preperitoneal fat thickness; MaxSFTupper, maximum subcutaneous fat thickness 2 cm above the umbilicus; MaxSFTlower, maximum subcutaneous fat thickness 2 cm underneath the umbilicus; MinSFT, minimum subcutaneous fat thickness. *One class of agreement gap between r. y Two classes of agreement gap between r.
0.87* 0.94 0.96 0.95 0.94* 0.92* 0.93 0.95 0.96 0.98* 0.86* 0.77y 0.96 0.98 0.86y 0.91* 0.92y 0.97 0.95 0.95y 0.97 0.96 0.96 0.97 0.95 0.89y 0.86 0.91 0.94 0.87y MinSFT MaxPFT MaxSFTupper MaxSFTlower IFT
0.91* 0.92y 0.87* 0.95 0.90
0.84* 0.79y 0.93* 0.97 0.90
0.91* 0.93* 0.97 0.98 0.98*
0.97* 0.98* 0.96 0.96 0.94*
0.95y 0.87 0.94 0.94 0.96y
0.95 0.95 0.96 0.99 0.96
Equipment 2 Equipment 1 Equipment 2 Equipment 1 Inspiration Expiration Inspiration Expiration Postmeal Premeal Postmeal
Operator 1 Operator 2 Operator 1
Premeal
Operator 2
Operator 1
Equipment Respiration Abdominal Distension
TABLE 4. Repeatability of Sonographic Measurements in Group A (Abdominal Distension), Group B (Respiration), and Group C (Equipment)
Operator 2
Academic Radiology, Vol 20, No 10, October 2013
ADIPOSITY: ISSUES FOR ULTRASOUND IMAGING
noninvasiveness, good tolerability, fast and easy use, wide availability and low costs—represent good reasons to encourage US in the clinical arena of fat quantification. In recent years, several studies have reported good correlations between US and gold standard techniques (ie, computed tomography and magnetic resonance imaging) in the assessment of abdominal adiposity, and between US adiposity indexes and different anthropometric and clinical indexes (2–11). On the other hand, clinical and laboratory parameters as well as metabolic and cardiovascular risk factors have also been investigated in light of US data (22–32). However, the reliability of US, in terms of accuracy and inter- and intra-observer agreement, remains a matter of debate, with performance values varying depending on studies (2–11,33). The heterogeneity of results might be imputed to poor standardization of US methods and research protocols and to a lack of analysis of US limitations in this field. Only a few authors have previously mentioned the potential limitations of US techniques for adiposity assessment, or have collaterally evaluated specific points or criticisms (2,31). However, to our knowledge, no one has researched the analysis of US limitations. Our aim was to perform a comprehensive analysis of such potential limitations and to provide limits of confidence and the spectrum of variability that might eventually be expected while performing an abdominal US examination for this purpose. In the field of body composition, US plays a role entirely devoted to tissue quantification. Therefore, an error in the measurement may affect US diagnostic results in the evaluation of fat thickness even more than in other more conventional aims of US examination. Potential limitations associated with US imaging can be divided into three main categories: intrinsic or technique limitations, patient-related limitations, and methodological limitations. From a different point of view, some of these limitations may be considered avoidable and some unavoidable. A few limitations are characteristic of US because of its physics and technology. US provides thicknesses of fat layers in abdominal sections, defined by linear measurements and distances between different ultrasound interfaces, representing the boundaries of different anatomic compartments. Therefore, some US parameters are not truly representative of adipose tissue volume per se and are not represented only by adipose tissue (31). For instance, MinSFT or MaxSFT and IFTare only thicknesses and not volumes of the compartments they represent. Furthermore, if the compartment for subcutaneous measures may be considered completely filled by adipose tissue (subcutaneous), the visceral compartment of IFT should not be. The IFT includes gut and other visceral organs with their complex anatomy, their movements, and their own filling. Technological limitations are also related to the US equipment. The different power of US machines may play a major role. The acquisition of technically challenging parameters, such as those representing visceral fat, might be more difficult 1283
BAZZOCCHI ET AL
with low-performing US equipments because of poor point of reference visualization without altering the normal shape of abdominal compartment (ie, exerting excessive pressure on the abdominal surface while scanning). However, in our study, the older and lower performing machine was well-correlated with the newer one concerning single measurements, and the use of the former was only associated with a slight decrease in reproducibility for visceral fat depots. Patient-related limitations are unavoidable, and when a patient is submitted for US examination, abdominal masses, ascites, large scars from previous abdominal surgery, and other features potentially affecting the measurements or quality of the diagnostic procedure should be taken into consideration, as should his or her BMI. A few investigations reported low accuracy for postpartum adolescents and very obese patients (2,33). The lack of standardization or wrong procedures in performing US produces methodologically critical states and not-comparable results. The operator’s training, the pressure of the probe on the skin, the positioning and respiratory condition of the patient, and the fasting period before the examination, if not considered, could lead to a decrease in accuracy and inter- and intra-operator agreement. We analyzed the articles reported in literature investigating the role of US in assessing visceral adiposity and found a high heterogeneity in the methods used, particularly for the breathing and fasting period of the patients (6,8,11,24– 28,31,32,34–36). In many cases, these conditions were even not reported (Table 1). Our study demonstrated significant changes in visceral fat measurements before and after a meal or during inspiration and expiration, altering fat thicknesses up to 73%: fasting state (high influence), breathing (moderate influence), and machine (mild influence, not statistically significant) (Table 2). The maximum changes observed in the values of adiposity were for visceral fat and for group A, in which IFT significantly increased after food consumption. This study confirms the results achieved in our first work investigating general reliability of US; all subcutaneous and visceral fat parameters showed good intra- and interoperator agreement. However, we proved that reproducibility and repeatability values, especially for visceral fat, are more stable in some conditions (fasting state, expiration, newer machine). It is easy to understand how the lack of a straight, standardized protocol could lead to a decrease in accuracy and reliability of US measurements of visceral adiposity, particularly when a longitudinal survey of fat indexes is needed for follow-up and control during diet regimens or other treatment for obesity and other metabolic-related diseases. The decision to test the three technical parameters (or conditions) in three different groups of patients was taken mainly to avoid a significant increase in time spent on a single US examination. A longer time per exam would have led to potentially altered conditions for both patient (compliance, breathing) and operator. 1284
Academic Radiology, Vol 20, No 10, October 2013
There are several limits to this study. First, the enrolled population included subjects presenting BMI within a range of normality or overweight and was generally young and healthy. Thus, results and conclusions for obese patients might be different (however, the differences between the two conditions might be expected even more pronounced, at least for group B [different ventilation mechanism] and C [trouble in reaching the deepest reference points]). Second, group A aimed at evaluating the effect of abdominal distension and food consumption on US measurements; however, abdominal distension should be considered according to the time waited after the meal (45 minutes) because this implies a different distribution and progression of the bolus along the gastrointestinal tract. The distension obtained 45 minutes after the meal in group A is predominantly epigastric (MaxPFT +, IFT ), whereas maximum bowel distension should be expected 1 to 2 hours after the meal. Thus, in a more advanced time point of the digestion process, IFT might be expected even more altered. Mild expiration or inspiration, with all the fickleness implicit within the adjective ‘‘mild,’’ was used because studies in the literature apply to this condition. This may help patients be more compliant with the examination, and the measurements more accurate and adherent to those of gold standard techniques. Mesenteric fat thickness was not considered in this study because of the poor accuracy achieved in previous examinations. Although this point may be open to criticisms, the choice was made for a simple reason: the aim of our previous work and of the present one is to test US for parameters of established clinical meaning and importance that may be transferred to fast clinical practice. Thus, if we failed in proposing a simplified and rapid method to assess mesenteric fat thickness, demonstrating its low accuracy (9), the aim of our future works will be directed in finding new and different parameters of visceral adiposity, according to their metabolic function and US assessability, but respecting the principle of fast and easy transfer from research to clinical practice. Our definitive aim is not (only) to investigate common parameters used by ultraspecialist clinicians who spent a lot of time to become skilled ‘‘fat measurers,’’ but to find how to make US as easy as possible and as significant as possible in the assessment of fat depots. The MaxPFT in this work was considered as a site of visceral fat, although this point might be also fairly but partially contested (2). Finally, in our study, we put US reliability for inter- and intra-observer agreements under investigation, showing the range of potential discrepancies in different examination conditions. However, accuracy should be also tested, as in our previous study. Furthermore, the last step to be analyzed remains the longitudinal stability of US measurements. The analysis of body composition is essential in the study of metabolism. Imaging techniques provide fundamental markers of fat composition and distribution. As the knowledge of adipose tissue physiopathology expands, more of these
Academic Radiology, Vol 20, No 10, October 2013
biomarkers are considered irreplaceable. US is becoming popular for the assessment of adiposity, but no one had studied this tool in light of its potential limitations. Errors in tissue quantification should be considered particularly important in this field and should be experienced and controlled according to technical and methodological issues. In writing this article, expect it to be useful to those specifically involved in the study of metabolic diseases as well as to those who approach the US for the clinical assessment of adiposity, enlightening important differences in using different methods.
ACKNOWLEDGMENT The authors would like to thank Dott. Emiliano Mariani, Medical Illustrator, (Faenza, Italy), author of Figure 1. REFERENCES 1. Heymsfield SB, Baumgartner RM, Allison DB, et al. Evaluation of total and regional adiposity. In: Bray GA, Bouchard C, eds. Handbook of obesity: clinical applications. New York: Marcel Dekker, 2005; 33–79. 2. Vlachos IS, Hatziioannou A, Perelas A, et al. Sonographic assessment of regional adiposity. AJR Am J Roentgenol 2007; 189:1545–1553. 3. Armellini F, Zamboni M, Rigo L, et al. The contribution of sonography to the measurement of intra-abdominal fat. J Clin Ultrasound 1990; 18:563–567. 4. Armellini F, Zamboni M, Robbi R, et al. Total and intra-abdominal fat measurements by ultrasound and computerized tomography. Int J Obes Relat Metab Disord 1993; 17:209–214. 5. Tornaghi G, Raiteri R, Pozzato C, et al. Anthropometric or ultrasonic measurements in assessment of visceral fat? A comparative study. Int J Obes Relat Metab Disord 1994; 18:771–775. 6. Stolk RP, Wink O, Zelissen PM, et al. Validity and reproducibility of ultrasonography for the measurement of intra-abdominal adipose tissue. Int J Obes Relat Metab Disord 2001; 25:1346–1351. 7. Hirooka M, Kumagi T, Kurose K, et al. A technique for the measurement of visceral fat by ultrasonography: comparison of measurements by ultrasonography and computed tomography. Intern Med 2005; 44:794–799. 8. Gradmark AM, Rydh A, Renstrom F, et al. Computed tomography-based validation of abdominal adiposity measurements from ultrasonography, dual-energy x-ray absorptiometry and anthropometry. Br J Nutr 2010; 104:582–588. 9. Bazzocchi A, Filonzi G, Ponti F, et al. Accuracy, reproducibility and repeatability of ultrasonography in the assessment of abdominal adiposity. Acad Radiol 2011; 18:1133–1143. 10. Abe T, Kawakami Y, Sugita M, et al. Use of B-mode ultrasound for visceral fat mass evaluation: comparisons with magnetic resonance imaging. Appl Human Sci 1995; 14:133–139. 11. Emmons RR, Garber CE, Cirnigliaro CM, et al. Assessment of measures for abdominal adiposity in persons with spinal cord injury. Ultrasound Med Biol 2011; 37:734–741. 12. Despres JP, Lemieux S, Lamarche B, et al. The insulin resistance dyslipidemic syndrome: contribution of visceral obesity and therapeutic implications. Int J Obes Relat Metab Disord 1995; 19(Suppl 1):S76–S86. 13. Wajchenberg BL. Subcutaneous and visceral adipose tissue: their relation to the metabolic syndrome. Endocr Rev 2000; 21:697–738. 14. Fox CS, Massaro JM, Hoffmann U, et al. Abdominal visceral and subcutaneous adipose tissue compartments: association with metabolic risk factors in the Framingham Heart Study. Circulation 2007; 116:39–48. 15. Bazzocchi A, Diano D, Battista G. How fat is fat? Lancet 2012; 380:e1.
ADIPOSITY: ISSUES FOR ULTRASOUND IMAGING
16. Item F, Konrad D. Visceral fat and metabolic inflammation: the portal theory revisited. Obes Rev 2012; 13(Suppl 2):S30–S39. 17. Lazar MA. How obesity causes diabetes: not a tall tale. Science 2005; 307: 373–375. 18. Suzuki R, Watanabe S, Hirai Y, et al. Abdominal wall fat index, estimated by ultrasonography, for assessment of the ratio of visceral fat to subcutaneous fat in the abdomen. Am J Med 1993; 95:309–314. 19. Soyama A, Nishikawa T, Ishizuka T, et al. Clinical usefulness of the thickness of preperitoneal and subcutaneous fat layer in the abdomen estimated by ultrasonography for diagnosing abdominal obesity in each type of impaired glucose tolerance in man. Endocr J 2005; 52: 229–236. 20. Taksali SE, Caprio S, Dziura J, et al. High visceral and low abdominal subcutaneous fat stores in the obese adolescent: a determinant of an adverse metabolic phenotype. Diabetes 2008; 57:367–371. 21. Lin LI. A concordance correlation coefficient to evaluate reproducibility. Biometrics 1989; 45:255–268. 22. Liu KH, Chan YL, Chan WB, et al. Sonographic measurement of mesenteric fat thickness is a good correlate with cardiovascular risk factors: comparison with subcutaneous and preperitoneal fat thickness, magnetic resonance imaging and anthropometric indexes. Int J Obes Relat Metab Disord 2003; 27:1267–1273. 23. Liu KH, Chan YL, Chan JC, et al. Association of carotid intima-media thickness with mesenteric, preperitoneal and subcutaneous fat thickness. Atherosclerosis 2005; 179:299–304. 24. Kim SK, Kim HJ, Hur KY, et al. Visceral fat thickness measured by ultrasonography can estimate not only visceral obesity but also risks of cardiovascular and metabolic diseases. Am J Clin Nutr 2004; 79: 593–599. 25. Ribeiro-Filho FF, Faria AN, Azjen S, et al. Methods of estimation of visceral fat: advantages of ultrasonography. Obes Res 2003; 11:1488–1494. 26. Ribeiro-Filho FF, Faria AN, Kohlmann O, Jr, et al. Ultrasonography for the evaluation of visceral fat and cardiovascular risk. Hypertension 2001; 38: 713–717. 27. Leite CC, Wajchenberg BL, Radominski R, et al. Intra-abdominal thickness by ultrasonography to predict risk factors for cardiovascular disease and its correlation with anthropometric measurements. Metabolism 2002; 51: 1034–1040. 28. De Lucia Rolfe E, Sleigh A, Finucane FM, et al. Ultrasound measurements of visceral and subcutaneous abdominal thickness to predict abdominal adiposity among older men and women. Obesity (Silver Spring) 2010; 18:625–631. 29. Merino-Ibarra E, Artieda M, Cenarro A, et al. Ultrasonography for the evaluation of visceral fat and the metabolic syndrome. Metabolism 2005; 54: 1230–1235. 30. Stolk RP, Meijer R, Mali WP, et al. Secondary Manifestations of Arterial Disease Study Group. Ultrasound measurements of intra-abdominal fat estimate the metabolic syndrome better than do measurements of waist circumference. Am J Clin Nutr 2003; 77:857–860. lez-Gonza lez NL, et al. Ultrasound eval31. Bartha JL, Marın-Segura P, Gonza uation of visceral fat and metabolic risk factors during early pregnancy. Obesity (Silver Spring) 2007; 15:2233–2239. 32. Oh J, Kim SK, Shin DK, et al. A simple ultrasound correlate of visceral fat. Ultrasound Med Biol 2011; 37:1444–1451. 33. Stevens-Simon C, Thureen P, Stamm E, et al. A comparison of four techniques for measuring central adiposity in postpartum adolescents. J Matern Fetal Med 2001; 10:209–213. 34. Koda M, Senda M, Kamba M, et al. Sonographic subcutaneous and visceral fat indices represent the distribution of body fat volume. Abdom Imaging 2007; 32:387–392. 35. Guldiken S, Tuncbilek N, Okten OO, et al. Visceral fat thickness determined using ultrasonography is associated with anthropometric and clinical parameters of metabolic syndrome. Int J Clin Pract 2006; 60: 1576–1581. 36. Sabir N, Sermez Y, Kazil S, et al. Correlation of abdominal fat accumulation and liver steatosis: importance of ultrasonographic and anthropometric measurements. Eur J Ultrasound 2001; 14:121–128.
1285
A
dipose tissue is one of the most enigmatic components of the body. Evaluation and quantification of adipose tissue are fundamental in the field of body composition as well as body composition is crucial for understanding human and animal metabolism and its alterations (1). Imaging has been essential in bringing body composition analysis to a clinical ground. In clinical practice, one of the most used techniques to assess the organ-tissue level of body composition is ultrasonography (US). This technique has demonstrated satisfying results in terms of accuracy and reproducibility (2–11). Moreover, US is ‘‘friendly’’ and oriented to one of the key points in the assessment of fat tissue: its distribution between subcutaneous and visceral compartments. The importance of this point has already been proven (12–17). Risk factors
Acad Radiol 2013; 20:1278–1285 From the Department of Specialized, Diagnostic, and Experimental Medicine, University of Bologna, Sant’Orsola - Malpighi Hospital, Via G. Massarenti 9, 40138 Bologna, Italy (A.B., G.F., F.P., M.A., C.S., E.S., G.B.); and Diagnostic and Interventional Radiology, ‘‘Rizzoli’’ Orthopaedic Institute, Bologna, Italy (A.B., U.A.). Received April 14, 2013; accepted July 24, 2013. Address correspondence to: A.B. e-mail: [email protected] ªAUR, 2013 http://dx.doi.org/10.1016/j.acra.2013.07.009
1278
for complications associated with obesity and the wider field of ‘‘metabolic diseases’’ start from fat, whether it is visceral or not. Recently, some studies underlined that the ratio between subcutaneous and visceral fat could be an important index of risk for impaired glucose tolerance, hypertriglyceridemia, hypercholesterolemia, dyslipidemia, fatty liver, and liver dysfunction (18,19), and adolescents with a prevalent visceral distribution were found to have a five times higher risk of developing metabolic syndrome than those who had a subcutaneous distribution (20). Although few works previously proposed (2–11), validated, and ‘‘consecrated’’ US for fat evaluation of different body sites, mainly of the abdomen, no one considered investigating US for important technical and methodological concerns. How might abdominal distension influence US assessment of adiposity? How might thoracic and abdominal movements during respiration affect the measurements? Does the use of older machines instead of new ones affect results? None of these points have yet been considered or defined. Because US, in this field, is advocated not for the qualitative evaluation of a tissue or an organ, but for the quantification of a tissue or an organ, all of these issues may be even more important. An error in measurement may represent a huge problem when comparing different patients, or patients at
Academic Radiology, Vol 20, No 10, October 2013
different time points in their clinical follow-up. It might be also interesting to know when and how these errors may occur and the cause of the variations in the measurements. Furthermore, all previous studies investigating US in this field showed inhomogeneous criteria in the assessment of abdominal adiposity and many variations in methods (Table 1). The aim of our study was to investigate the impact of abdominal distension, patients’ breathing, and US equipment in the evaluation of subcutaneous and visceral abdominal fat by US. MATERIALS AND METHODS Study Design and Population
Ninety healthy volunteers (45 males and 45 females) were consecutively enrolled in the study. Exclusion criteria were: history of pathologies involving the trunk (thorax and abdomen) and potentially affecting its anatomy, including past surgical treatments or implanted devices. The patients were randomly assigned to three groups and equally distributed by sex (15 males and 15 females for each group). Subjects in group A (32.9 5.4 years old; body mass index [BMI] 24.1 2.7 kg/m2, range 21.2–28.9) were submitted to US examination (Esaote Technos MPX-2003; Esaote, Genoa, Italy) before (in a fasting state of at least 12 hours) and 45 minutes after a standardized meal (lunch), with measurements performed during a mild expiration. Subjects of group B (29.8 3.8 years old; BMI 23.6 2.9 kg/m2, range 18.4–28.2) underwent US assessment in mild expiration and 30 minutes later in mild inspiration, both times in a fasting state (at least 12 hours) and with the same US machine as for group A. Group C (33.5 4.1 years old; BMI 25.3 2.6 kg/m2, range 20.4–29.7) was evaluated in a fasting state (at least 12 hours) and during a mild expiration by using US equipments of different generations: equipment 1 (Esaote Technos MPX; year 2003; tissue harmonic imaging, yes; compounding, no; maximum scan depth, 28.5 cm; gray levels, 256) and equipment 2 (Hitachi EUB 525 eidos; Hitachi Denshi, Tokyo, Japan; year, 1998; tissue harmonic imaging, no; compounding, no; maximum scan depth, 24 cm; gray levels, 256). Thus, patients of all groups were evaluated in two different conditions: (a) fasting state, (b) breathing, and (c) equipment. Moreover these conditions were all tested by two skilled radiologists twice to evaluate inter- and intra-observer variability of the measurements associated with the specified conditions. The ‘‘state-of-the-art’’ evaluation (patients in group C, equipment 1) was also repeated twice, with the lowest and the highest hand (probe) pressure on the abdomen to verify the potential influence of this element. To blind the study, all measurements were independently collected by a third physician, and the readers were blinded to each others in the US scans. The study was carried out respecting the principles of the Declaration of Helsinki. The study protocol was approved by the local institutional
ADIPOSITY: ISSUES FOR ULTRASOUND IMAGING
review board and all participants signed an informed consent form. Ultrasound Methods
All measurements were performed with subjects in a supine position with arms at sides. Five parameters, representative of subcutaneous and visceral fat, were considered in the three groups: 1) minimum subcutaneous fat thickness (MinSFT); maximum subcutaneous fat thickness, upper (MaxSFTupper); and maximum subcutaneous fat thickness, lower (MaxSFTlower); 2) maximum preperitoneal fat thickness (MaxPFT); and 3) intra-abdominal fat thickness (IFT). MaxPFT measurement was performed just below the xiphoid process in the epigastric region, with a longitudinal scan on the xiphoumbilical line, as the major distance between the anterior surface of the peritoneum covering the liver (left lobe) and the posterior surface of linea alba. MinSFT was measured in the same anatomic region as the maximum preperitoneal fat. MaxSFT was assessed at two different sites on the linea alba, 2 cm above and 2 cm underneath the umbilicus (MaxSFTupper and MaxSFTlower, respectively), with a longitudinal scan. MinSFT, MaxSFTupper, and MaxSFTlower were defined as the distance between the anterior surface of linea alba and the fat-skin barrier. IFT was measured as the distance between the anterior wall of the aorta and the posterior surface of linea alba, 2 cm above the umbilicus, with a transverse scan (Fig 1). MaxPFT, MinSFT, MaxSFTupper, and MaxSFTlower were measured using a linear probe (7.5 MHz) kept perpendicular to the skin and with hand pressure on the abdomen as light as possible in order not to compress the fat layers (with the exception of the circumstance in group C); IFT was assessed using a convex probe (3.5 MHz) with the same modalities. All of these parameters were investigated by our team in 2011 in terms of accuracy, inter-observer, and intraobserver agreements in people with normal and high BMI (9). Statistical Analysis
The normal distribution of our sample population was tested by skewness and excess kurtosis test. Normal ranges were considered for values between 2 and +2. Results are reported as mean and standard deviation (SD), and integrated with Delta (D) and Delta percentage (D%). Student’s t-test was used to analyze the statistical impact of abdominal distension, respiration, and equipment on US abdominal adiposity measurements; P was considered significant for values less than .05. Lin’s correlation coefficient (r) was used to assess intra- and inter-operator agreements in all groups for all conditions and parameters. A poor correlation was assumed for r values less than 0.70; moderate correlation with r between 0.70 and 0.79; good correlation with r between 0.80 and 0.89; very good correlation between 0.90 and 0.94; and excellent correlation with r of 0.95 or more (21). A ‘‘significant’’ difference in reproducibility and 1279
BAZZOCCHI ET AL
Academic Radiology, Vol 20, No 10, October 2013
TABLE 1. Methods and Technical Issues of the Principal Articles Measuring Visceral Fat Thickness with US Between 2001 and 2011
Author Publication
Fasting Condition
Breath Condition
Emmons RR, Ultrasound Med Biol (2011)
Not declared
Upon exhalation
Oh J, Ultrasound Med Biol (2011)
After fasting at least for 4 hours
Immediately After expiration
Anatomical Landmarks of Visceral Fat From the linea alba to the anterior aorta
From the anterior wall of the aorta to the internal face of the rectus abdominis muscle De Lucia Rolfe E, Refrain from eating At the end of a quiet From the peritoneal boundary Obesity (2010) 10 hours before expiration to the corpus of the lumbar vertebra Gradmark AM, Br J Fasting state Hold breath during From the inside of the bowel Nutr (2010) mid-inspiration wall to the spine Bartha JL, Obesity Fasting state Immediately at the From the anterior wall of the (2007) end of expiration aorta to the internal layer of the rectoabdominal muscle Koda M, Abdom Fasting overnight Not declared From the internal face of the Imaging (2007) rectus abdominis muscle (linea alba) to the anterior wall of the vertebra at the level of the umbilicus Not declared Breath-hold From the internal face of the Guldiken S, Int J Clin Pract (2006) abdominal muscle to the anterior wall of the aorta Kim SK, Am J Clin Not declared Immediately after From the anterior wall of the Nutr (2004) respiration aorta to the internal face of the rectoabdominal muscle Ribeiro-Filho FF, Fasting state Not declared From the internal face of Obes Res (2003) rectoabdominal muscle to the anterior wall of the aorta Leite CC, Metabolism Not declared Not declared From the internal face of the (2002) abdominal muscles to the posterior wall of the aorta Stolk RP, Int J Obes Relat Not declared At the end of a quiet From the posterior edge of the Metab Disord (2001) expiration abdominal muscles to the lumbar spine or psoas muscles Sabir N, Eur J Ultrasound Not declared Breath-hold From the internal face of the (2001) abdominal muscle to the anterior wall of the aorta.
repeatability between the measurements was defined for results (r) belonging to different classes of agreements (poor, moderate, etc.). Statistical analysis was performed using MedCalc version 11.4.2 (MedCalc Software, Mariakerke, Belgium).
US Equipment GE Logiq Book XP (GE Healthcare, Buckinghamshire, UK) Prosound a10 (Aloka, Tokyo, Japan) Logic Book XP ultrasound (GE Healthcare, Bedford, UK) Acuson Sequoia, (Siemens, Mountain Woods, California) Esaote Technos MPX (Esaote S.p.A., Genova, Italy) Aloka SSD-2000 (Aloka, Tokyo, Japan)
Sonoline Elegra, (Siemens Medical Solutions, Inc., Issaquah, Washington) SA 9900 (Medison, Seoul, Korea) Not declared
Toshiba Sonolayer SSA-250 A (Otawara Shi, Tochigiken, Japan) ATL HDI 3000 system (Bothell, Washington)
GE logiq 200
% = 24.1 21.3, absolute value D% = 25.7 21.2, minimum D% = 13.3, maximum D% = 73.7, P < .0001) (Fig 2). Subcutaneous measures and MaxPFT were not affected by abdominal distension, at least not with a statistical significance (Fig 3).
US values and differences for technical conditions are presented in Table 2.
Group B. As for group A, IFT showed the most important variation (D 5.6 8.2 mm, D% = 9.2 20.4, absolute value D% = 18.75 11.91, minimum D% = 31.4, maximum D% = 42.6, P < .0001). On the other side, none of the other parameters showed significant variation between the two different conditions.
Group A. IFT showed important variations before and after the meal, from 47.4 17.3 mm to 56.9 15.2 mm (D
Group C. Notwithstanding the important technological gap between the two US machines, no significant difference
RESULTS Descriptive Statistics
1280
Academic Radiology, Vol 20, No 10, October 2013
Figure 1. Abdominal sagittal section showing anatomical landmarks and parameters of adiposity as evaluated by ultrasonography: minimum subcutaneous fat thickness (a); maximum preperitoneal fat thickness (b); maximum subcutaneous fat thickness, upper (c); intra-abdominal fat thickness (d); and maximum subcutaneous fat thickness, lower (e).
resulted in the measurements performed by the two machines. The level of tissue compression by the probe produced extremely relevant changes of values for all parameters (D% ranging between 20.6% and 47.7% for subcutaneous parameters and between 53.3% and 74.8% for visceral parameters).
Analysis of Reliability
All data are shown in Table 3 (reproducibility) and Table 4 (repeatability).
Reproducibility. All parameters showed moderate to excellent reproducibility in all conditions evaluated (r between 0.78 and 0.95). IFT, MaxPFT, and MinSFT were the only parameters affected by statistically relevant changes in reproducibility. IFT showed significant variations of interobserver agreement in all groups (r 0.92–0.85 in group A, 0.91–0.86 in group B, 0.95–0.87 in group C). MaxPFT and MinSFT were also critically unstable, but only for group B (r 0.83–0.78 and 0.92–0.86, respectively).
Repeatability. All measurements were reported with moderate to excellent values of repeatability, for both operators (r ranging between 0.77 and 0.99). However, IFTwas again affected by statistically considerable changes, although these were more dependent on single operators in groups A and B.
TABLE 2. Ultrasound Values, Delta (D), and Delta Percentage (D%) in Group A (Abdominal Distension), Group B (Respiration), and Group C (Equipment)
MinSFT MaxPFT MaxSFTupper MaxSFTlower IFT
Respiration
Equipment
Pre-m (mm) Mean SD min/max
Post-m (mm) D (mm) D% (%) Exp (mm) Mean SD Mean SD Mean SD Mean SD min/max min/max min/max min/max
Insp (mm) Mean SD min/max
D (mm) Mean SD min/max
D% (%) Mean SD min/max
E 1 (mm) E 2 (mm) D (mm) Mean SD Mean SD Mean SD min/max min/max min/max
7.6 3.4 3.7/14.6 14.3 4.2 5.3/22.4 18.8 7.2 6.9/31.4 19.7 9.1 5.6/40.8 47.4* 17.3 27.3/92.4
7.6 2.9 3.0/14.7 15.0 4.9 5.5/21.9 18.8 7.1 6.5/29.0 20.1 9.0 6.9/39.0 56.9* 15.2 35.9/95.6
5.8 2.4 2.7/10.3 13.6 3.9 7.9/21.2 16.2 6.2 6.2/26.8 17.6 7.6 6.5/33.1 43.6* 12.8 20.3/75.1
0.1 1.1 3.2/3.0 0.3 2.2 3.9/5.9 0.5 1.8 3.2/3.8 0.5 1.9 5.4/3.4 5.6 8.2 11.2/15.2
0.6 20.2 34.0/42.2 4.2 15.2 40.3/27.8 4.1 11.4 27.4/16.3 4.6 11.8 10.6/26.3 9.2 20.4 31.4/42.6
8.6 3.0 8.4 2.8 3.1/15.9 3.4/15.2 14.5 5.2 13.5 5.1 5.3/22.4 5.0/21.3 19.5 7.9 19.2 8.0 7.9/31.5 6.9/29.5 21.1 8.9 21.9 9.8 5.7/40.8 7.6/41.8 50.9 14.1 51.3 16.0 31.4/82.1 27.3/88.4
0.0 1.8 3.7/7.7 0.8 1.7 3.2/6.2 0.0 2.1 6.0/5.8 0.4 2.2 4.2/7.3 9.5 8.1 14.2/25.6
1.0 22.8 54.4/53.4 4.8 12.5 20.3/40.2 0.4 11.5 32.3/31.5 4.0 13.9 18.8/42.6 24.1 21.3 15.3/73.7
5.7 2.7 5.8/12.6 13.9 3.7 7.9/21.8 16.7 6.0 7.0/27.1 18.1 7.3 8.7/34.3 38.0* 9.9 21.9/59.1
0.1 0.9 2.2/2.0 1.0 2.4 10.5/1.1 0.2 2.5 8.7/3.5 0.8 2.9 7.7/7.3 0.4 4.1 8.6/9.6
D% (%) Mean SD min/max 3.6 11.4 30.5/13.4 10.2 20.4 68.2/5.6 2.2 16.5 65.9/16.6 6.0 15.6 37.0/34.1 0.1 10.2 27.3/26.0
1281
E 1, equipment 1; E 2, equipment 2; exp, expiration; IFT, intra-abdominal fat thickness; insp, inspiration; max, maximum; MaxPFT, maximum preperitoneal fat thickness; MaxSFTlower, maximum subcutaneous fat thickness 2 cm underneath the umbilicus; MaxSFTupper, maximum subcutaneous fat thickness 2 cm above the umbilicus; min, minimum; MinSFT, minimum subcutaneous fat thickness; pre-m, premeal; post-m, postmeal. *Student’s t-test: significant change with P < .005.
ADIPOSITY: ISSUES FOR ULTRASOUND IMAGING
Abdominal Distension
BAZZOCCHI ET AL
Academic Radiology, Vol 20, No 10, October 2013
Figure 2. Opposite conditions for intraabdominal fat thickness in three subjects, one for each group (a, b, and c); before (a1) and after (a2) food consumption; in expiration (b1) and inspiration (b2); and acquired by equipment 1 (c1) and equipment 2 (c2). Thicknesses are expressed in millimeters.
Figure 3. Opposite conditions in three subjects, one for each group (a, b, and c) for subcutaneous fat parameters and maximum preperitoneal fat thickness: minimum subcutaneous fat thickness and maximum preperitoneal fat thickness before (a1) and after (a2) food consumption; maximum subcutaneous fat thickness, upper, during expiration (b1) and during inspiration (b2); and maximum fat thickness, lower, with equipment 1 (c1) and equipment 2 (c2). Thicknesses are expressed in millimeters.
TABLE 3. Reproducibility of Sonographic Measurements in Group A (Abdominal Distension), Group B (Respiration), and Group C (Equipment) Abdominal Distension
MinSFT MaxPFT MaxSFTupper MaxSFTlower IFT
Respiration
Equipment
Premeal
Postmeal
Expiration
Inspiration
Equipment 1
Equipment 2
0.86 0.82 0.90 0.91 0.92*
0.83 0.85 0.94 0.93 0.85*
0.92* 0.83* 0.95 0.94 0.91*
0.86* 0.78* 0.95 0.91 0.86*
0.87 0.84 0.94 0.92 0.95y
0.81 0.81 0.93 0.94 0.87y
IFT, intra-abdominal fat thickness; MaxPFT, maximum preperitoneal fat thickness; MaxSFTupper, maximum subcutaneous fat thickness 2 cm above the umbilicus; MaxSFTlower, maximum subcutaneous fat thickness 2 cm underneath the umbilicus; MinSFT, minimum subcutaneous fat thickness. *One class of agreement gap between r. y Two classes of agreement gap between r.
MinSFT showed significant variations in all groups, whereas MaxPFT showed important variations in group A and group C. All other parameters suffered from few variations, occasional variations, or no variation at all. The compression of the abdomen did not statistically influence the level of reproducibility and repeatability. 1282
DISCUSSION Today, US is one of the most suitable tools for evaluating abdominal adiposity in clinical practice, especially in the differential assessment of subcutaneous and visceral adipose depots. Some of the most attractive features of US—such as
IFT, intra-abdominal fat thickness; MaxPFT, maximum preperitoneal fat thickness; MaxSFTupper, maximum subcutaneous fat thickness 2 cm above the umbilicus; MaxSFTlower, maximum subcutaneous fat thickness 2 cm underneath the umbilicus; MinSFT, minimum subcutaneous fat thickness. *One class of agreement gap between r. y Two classes of agreement gap between r.
0.87* 0.94 0.96 0.95 0.94* 0.92* 0.93 0.95 0.96 0.98* 0.86* 0.77y 0.96 0.98 0.86y 0.91* 0.92y 0.97 0.95 0.95y 0.97 0.96 0.96 0.97 0.95 0.89y 0.86 0.91 0.94 0.87y MinSFT MaxPFT MaxSFTupper MaxSFTlower IFT
0.91* 0.92y 0.87* 0.95 0.90
0.84* 0.79y 0.93* 0.97 0.90
0.91* 0.93* 0.97 0.98 0.98*
0.97* 0.98* 0.96 0.96 0.94*
0.95y 0.87 0.94 0.94 0.96y
0.95 0.95 0.96 0.99 0.96
Equipment 2 Equipment 1 Equipment 2 Equipment 1 Inspiration Expiration Inspiration Expiration Postmeal Premeal Postmeal
Operator 1 Operator 2 Operator 1
Premeal
Operator 2
Operator 1
Equipment Respiration Abdominal Distension
TABLE 4. Repeatability of Sonographic Measurements in Group A (Abdominal Distension), Group B (Respiration), and Group C (Equipment)
Operator 2
Academic Radiology, Vol 20, No 10, October 2013
ADIPOSITY: ISSUES FOR ULTRASOUND IMAGING
noninvasiveness, good tolerability, fast and easy use, wide availability and low costs—represent good reasons to encourage US in the clinical arena of fat quantification. In recent years, several studies have reported good correlations between US and gold standard techniques (ie, computed tomography and magnetic resonance imaging) in the assessment of abdominal adiposity, and between US adiposity indexes and different anthropometric and clinical indexes (2–11). On the other hand, clinical and laboratory parameters as well as metabolic and cardiovascular risk factors have also been investigated in light of US data (22–32). However, the reliability of US, in terms of accuracy and inter- and intra-observer agreement, remains a matter of debate, with performance values varying depending on studies (2–11,33). The heterogeneity of results might be imputed to poor standardization of US methods and research protocols and to a lack of analysis of US limitations in this field. Only a few authors have previously mentioned the potential limitations of US techniques for adiposity assessment, or have collaterally evaluated specific points or criticisms (2,31). However, to our knowledge, no one has researched the analysis of US limitations. Our aim was to perform a comprehensive analysis of such potential limitations and to provide limits of confidence and the spectrum of variability that might eventually be expected while performing an abdominal US examination for this purpose. In the field of body composition, US plays a role entirely devoted to tissue quantification. Therefore, an error in the measurement may affect US diagnostic results in the evaluation of fat thickness even more than in other more conventional aims of US examination. Potential limitations associated with US imaging can be divided into three main categories: intrinsic or technique limitations, patient-related limitations, and methodological limitations. From a different point of view, some of these limitations may be considered avoidable and some unavoidable. A few limitations are characteristic of US because of its physics and technology. US provides thicknesses of fat layers in abdominal sections, defined by linear measurements and distances between different ultrasound interfaces, representing the boundaries of different anatomic compartments. Therefore, some US parameters are not truly representative of adipose tissue volume per se and are not represented only by adipose tissue (31). For instance, MinSFT or MaxSFT and IFTare only thicknesses and not volumes of the compartments they represent. Furthermore, if the compartment for subcutaneous measures may be considered completely filled by adipose tissue (subcutaneous), the visceral compartment of IFT should not be. The IFT includes gut and other visceral organs with their complex anatomy, their movements, and their own filling. Technological limitations are also related to the US equipment. The different power of US machines may play a major role. The acquisition of technically challenging parameters, such as those representing visceral fat, might be more difficult 1283
BAZZOCCHI ET AL
with low-performing US equipments because of poor point of reference visualization without altering the normal shape of abdominal compartment (ie, exerting excessive pressure on the abdominal surface while scanning). However, in our study, the older and lower performing machine was well-correlated with the newer one concerning single measurements, and the use of the former was only associated with a slight decrease in reproducibility for visceral fat depots. Patient-related limitations are unavoidable, and when a patient is submitted for US examination, abdominal masses, ascites, large scars from previous abdominal surgery, and other features potentially affecting the measurements or quality of the diagnostic procedure should be taken into consideration, as should his or her BMI. A few investigations reported low accuracy for postpartum adolescents and very obese patients (2,33). The lack of standardization or wrong procedures in performing US produces methodologically critical states and not-comparable results. The operator’s training, the pressure of the probe on the skin, the positioning and respiratory condition of the patient, and the fasting period before the examination, if not considered, could lead to a decrease in accuracy and inter- and intra-operator agreement. We analyzed the articles reported in literature investigating the role of US in assessing visceral adiposity and found a high heterogeneity in the methods used, particularly for the breathing and fasting period of the patients (6,8,11,24– 28,31,32,34–36). In many cases, these conditions were even not reported (Table 1). Our study demonstrated significant changes in visceral fat measurements before and after a meal or during inspiration and expiration, altering fat thicknesses up to 73%: fasting state (high influence), breathing (moderate influence), and machine (mild influence, not statistically significant) (Table 2). The maximum changes observed in the values of adiposity were for visceral fat and for group A, in which IFT significantly increased after food consumption. This study confirms the results achieved in our first work investigating general reliability of US; all subcutaneous and visceral fat parameters showed good intra- and interoperator agreement. However, we proved that reproducibility and repeatability values, especially for visceral fat, are more stable in some conditions (fasting state, expiration, newer machine). It is easy to understand how the lack of a straight, standardized protocol could lead to a decrease in accuracy and reliability of US measurements of visceral adiposity, particularly when a longitudinal survey of fat indexes is needed for follow-up and control during diet regimens or other treatment for obesity and other metabolic-related diseases. The decision to test the three technical parameters (or conditions) in three different groups of patients was taken mainly to avoid a significant increase in time spent on a single US examination. A longer time per exam would have led to potentially altered conditions for both patient (compliance, breathing) and operator. 1284
Academic Radiology, Vol 20, No 10, October 2013
There are several limits to this study. First, the enrolled population included subjects presenting BMI within a range of normality or overweight and was generally young and healthy. Thus, results and conclusions for obese patients might be different (however, the differences between the two conditions might be expected even more pronounced, at least for group B [different ventilation mechanism] and C [trouble in reaching the deepest reference points]). Second, group A aimed at evaluating the effect of abdominal distension and food consumption on US measurements; however, abdominal distension should be considered according to the time waited after the meal (45 minutes) because this implies a different distribution and progression of the bolus along the gastrointestinal tract. The distension obtained 45 minutes after the meal in group A is predominantly epigastric (MaxPFT +, IFT ), whereas maximum bowel distension should be expected 1 to 2 hours after the meal. Thus, in a more advanced time point of the digestion process, IFT might be expected even more altered. Mild expiration or inspiration, with all the fickleness implicit within the adjective ‘‘mild,’’ was used because studies in the literature apply to this condition. This may help patients be more compliant with the examination, and the measurements more accurate and adherent to those of gold standard techniques. Mesenteric fat thickness was not considered in this study because of the poor accuracy achieved in previous examinations. Although this point may be open to criticisms, the choice was made for a simple reason: the aim of our previous work and of the present one is to test US for parameters of established clinical meaning and importance that may be transferred to fast clinical practice. Thus, if we failed in proposing a simplified and rapid method to assess mesenteric fat thickness, demonstrating its low accuracy (9), the aim of our future works will be directed in finding new and different parameters of visceral adiposity, according to their metabolic function and US assessability, but respecting the principle of fast and easy transfer from research to clinical practice. Our definitive aim is not (only) to investigate common parameters used by ultraspecialist clinicians who spent a lot of time to become skilled ‘‘fat measurers,’’ but to find how to make US as easy as possible and as significant as possible in the assessment of fat depots. The MaxPFT in this work was considered as a site of visceral fat, although this point might be also fairly but partially contested (2). Finally, in our study, we put US reliability for inter- and intra-observer agreements under investigation, showing the range of potential discrepancies in different examination conditions. However, accuracy should be also tested, as in our previous study. Furthermore, the last step to be analyzed remains the longitudinal stability of US measurements. The analysis of body composition is essential in the study of metabolism. Imaging techniques provide fundamental markers of fat composition and distribution. As the knowledge of adipose tissue physiopathology expands, more of these
Academic Radiology, Vol 20, No 10, October 2013
biomarkers are considered irreplaceable. US is becoming popular for the assessment of adiposity, but no one had studied this tool in light of its potential limitations. Errors in tissue quantification should be considered particularly important in this field and should be experienced and controlled according to technical and methodological issues. In writing this article, expect it to be useful to those specifically involved in the study of metabolic diseases as well as to those who approach the US for the clinical assessment of adiposity, enlightening important differences in using different methods.
ACKNOWLEDGMENT The authors would like to thank Dott. Emiliano Mariani, Medical Illustrator, (Faenza, Italy), author of Figure 1. REFERENCES 1. Heymsfield SB, Baumgartner RM, Allison DB, et al. Evaluation of total and regional adiposity. In: Bray GA, Bouchard C, eds. Handbook of obesity: clinical applications. New York: Marcel Dekker, 2005; 33–79. 2. Vlachos IS, Hatziioannou A, Perelas A, et al. Sonographic assessment of regional adiposity. AJR Am J Roentgenol 2007; 189:1545–1553. 3. Armellini F, Zamboni M, Rigo L, et al. The contribution of sonography to the measurement of intra-abdominal fat. J Clin Ultrasound 1990; 18:563–567. 4. Armellini F, Zamboni M, Robbi R, et al. Total and intra-abdominal fat measurements by ultrasound and computerized tomography. Int J Obes Relat Metab Disord 1993; 17:209–214. 5. Tornaghi G, Raiteri R, Pozzato C, et al. Anthropometric or ultrasonic measurements in assessment of visceral fat? A comparative study. Int J Obes Relat Metab Disord 1994; 18:771–775. 6. Stolk RP, Wink O, Zelissen PM, et al. Validity and reproducibility of ultrasonography for the measurement of intra-abdominal adipose tissue. Int J Obes Relat Metab Disord 2001; 25:1346–1351. 7. Hirooka M, Kumagi T, Kurose K, et al. A technique for the measurement of visceral fat by ultrasonography: comparison of measurements by ultrasonography and computed tomography. Intern Med 2005; 44:794–799. 8. Gradmark AM, Rydh A, Renstrom F, et al. Computed tomography-based validation of abdominal adiposity measurements from ultrasonography, dual-energy x-ray absorptiometry and anthropometry. Br J Nutr 2010; 104:582–588. 9. Bazzocchi A, Filonzi G, Ponti F, et al. Accuracy, reproducibility and repeatability of ultrasonography in the assessment of abdominal adiposity. Acad Radiol 2011; 18:1133–1143. 10. Abe T, Kawakami Y, Sugita M, et al. Use of B-mode ultrasound for visceral fat mass evaluation: comparisons with magnetic resonance imaging. Appl Human Sci 1995; 14:133–139. 11. Emmons RR, Garber CE, Cirnigliaro CM, et al. Assessment of measures for abdominal adiposity in persons with spinal cord injury. Ultrasound Med Biol 2011; 37:734–741. 12. Despres JP, Lemieux S, Lamarche B, et al. The insulin resistance dyslipidemic syndrome: contribution of visceral obesity and therapeutic implications. Int J Obes Relat Metab Disord 1995; 19(Suppl 1):S76–S86. 13. Wajchenberg BL. Subcutaneous and visceral adipose tissue: their relation to the metabolic syndrome. Endocr Rev 2000; 21:697–738. 14. Fox CS, Massaro JM, Hoffmann U, et al. Abdominal visceral and subcutaneous adipose tissue compartments: association with metabolic risk factors in the Framingham Heart Study. Circulation 2007; 116:39–48. 15. Bazzocchi A, Diano D, Battista G. How fat is fat? Lancet 2012; 380:e1.
ADIPOSITY: ISSUES FOR ULTRASOUND IMAGING
16. Item F, Konrad D. Visceral fat and metabolic inflammation: the portal theory revisited. Obes Rev 2012; 13(Suppl 2):S30–S39. 17. Lazar MA. How obesity causes diabetes: not a tall tale. Science 2005; 307: 373–375. 18. Suzuki R, Watanabe S, Hirai Y, et al. Abdominal wall fat index, estimated by ultrasonography, for assessment of the ratio of visceral fat to subcutaneous fat in the abdomen. Am J Med 1993; 95:309–314. 19. Soyama A, Nishikawa T, Ishizuka T, et al. Clinical usefulness of the thickness of preperitoneal and subcutaneous fat layer in the abdomen estimated by ultrasonography for diagnosing abdominal obesity in each type of impaired glucose tolerance in man. Endocr J 2005; 52: 229–236. 20. Taksali SE, Caprio S, Dziura J, et al. High visceral and low abdominal subcutaneous fat stores in the obese adolescent: a determinant of an adverse metabolic phenotype. Diabetes 2008; 57:367–371. 21. Lin LI. A concordance correlation coefficient to evaluate reproducibility. Biometrics 1989; 45:255–268. 22. Liu KH, Chan YL, Chan WB, et al. Sonographic measurement of mesenteric fat thickness is a good correlate with cardiovascular risk factors: comparison with subcutaneous and preperitoneal fat thickness, magnetic resonance imaging and anthropometric indexes. Int J Obes Relat Metab Disord 2003; 27:1267–1273. 23. Liu KH, Chan YL, Chan JC, et al. Association of carotid intima-media thickness with mesenteric, preperitoneal and subcutaneous fat thickness. Atherosclerosis 2005; 179:299–304. 24. Kim SK, Kim HJ, Hur KY, et al. Visceral fat thickness measured by ultrasonography can estimate not only visceral obesity but also risks of cardiovascular and metabolic diseases. Am J Clin Nutr 2004; 79: 593–599. 25. Ribeiro-Filho FF, Faria AN, Azjen S, et al. Methods of estimation of visceral fat: advantages of ultrasonography. Obes Res 2003; 11:1488–1494. 26. Ribeiro-Filho FF, Faria AN, Kohlmann O, Jr, et al. Ultrasonography for the evaluation of visceral fat and cardiovascular risk. Hypertension 2001; 38: 713–717. 27. Leite CC, Wajchenberg BL, Radominski R, et al. Intra-abdominal thickness by ultrasonography to predict risk factors for cardiovascular disease and its correlation with anthropometric measurements. Metabolism 2002; 51: 1034–1040. 28. De Lucia Rolfe E, Sleigh A, Finucane FM, et al. Ultrasound measurements of visceral and subcutaneous abdominal thickness to predict abdominal adiposity among older men and women. Obesity (Silver Spring) 2010; 18:625–631. 29. Merino-Ibarra E, Artieda M, Cenarro A, et al. Ultrasonography for the evaluation of visceral fat and the metabolic syndrome. Metabolism 2005; 54: 1230–1235. 30. Stolk RP, Meijer R, Mali WP, et al. Secondary Manifestations of Arterial Disease Study Group. Ultrasound measurements of intra-abdominal fat estimate the metabolic syndrome better than do measurements of waist circumference. Am J Clin Nutr 2003; 77:857–860. lez-Gonza lez NL, et al. Ultrasound eval31. Bartha JL, Marın-Segura P, Gonza uation of visceral fat and metabolic risk factors during early pregnancy. Obesity (Silver Spring) 2007; 15:2233–2239. 32. Oh J, Kim SK, Shin DK, et al. A simple ultrasound correlate of visceral fat. Ultrasound Med Biol 2011; 37:1444–1451. 33. Stevens-Simon C, Thureen P, Stamm E, et al. A comparison of four techniques for measuring central adiposity in postpartum adolescents. J Matern Fetal Med 2001; 10:209–213. 34. Koda M, Senda M, Kamba M, et al. Sonographic subcutaneous and visceral fat indices represent the distribution of body fat volume. Abdom Imaging 2007; 32:387–392. 35. Guldiken S, Tuncbilek N, Okten OO, et al. Visceral fat thickness determined using ultrasonography is associated with anthropometric and clinical parameters of metabolic syndrome. Int J Clin Pract 2006; 60: 1576–1581. 36. Sabir N, Sermez Y, Kazil S, et al. Correlation of abdominal fat accumulation and liver steatosis: importance of ultrasonographic and anthropometric measurements. Eur J Ultrasound 2001; 14:121–128.
1285