A Longitudinal Study to Assess Abdominal Adiposity by Sonography,DXA,and BMI

Abstract

Obesity has been linked to several chronic health conditions, including increased risk for cardiovascular disease, type 2 diabetes, some cancers, and metabolic syndrome. Understanding the trends in metabolic syndrome and examining those who are at an increased risk for metabolic conditions is paramount. Currently body mass index (BMI), waist circumference (WC), and dual-energy X-ray absorptiometry (DXA) are believed to be appropriate screening measures for assessing body composition and metabolic syndrome risk. This approach uses diagnostic medical sonography (DMS) as a tool for assessing abdominal adiposity in hopes of this methodology being adopted to help determine an individual’s risk for metabolic conditions. A convenience sample of overweight and obese cancer survivors provided several anthropometric measures as well as imaging data to determine if sonographic measures of abdominal fat could be assessed longitudinally. These measures were then compared to corresponding measures of BMI, WC, and DXA (android percent body fat, subscores). These measures demonstrated moderately positive association and were statistically significant. DMS is an indirect imaging tool that could be used in correlation with other measures to help determine an individual’s risk for metabolic syndrome.

Keywords

abdominal adiposity visceral fat subcutaneous fat sonographic measures sonography DXA BMI body composition

Approximately 93 million adults and over 13 million children and adolescents across the United States are considered overweight or obese.¹ Health care expenditures are shown to be 81% greater in obese adults versus those of normal weight.^2,3 The prevalence of type 2 diabetes and hypertension has steadily increased in the United States, with metabolic syndrome rates estimated at more than 30%.⁴ Due to the increased risk for metabolic conditions in America’s obese populations, identifying patients at risk with a reliable and practical tool is vital to providing preventive health measures. Given the link between abdominal adiposity and metabolic syndrome, determining an individual’s risk for disease by measuring abdominal adiposity is greatly needed.

The use of diagnostic medical sonography (DMS) has been shown to be a reliable and easily accessible tool when measuring abdominal adiposity.⁵ DMS is an indirect imaging tool that could be used in correlation with other measures to help determine an individual’s risk for metabolic syndrome. The value of DMS for gauging abdominal obesity and the risk for metabolic disease is that it is relatively inexpensive, noninvasive, and a validated indirect assessment tool. DMS has the ability to measure increased abdominal adiposity and monitor changes in body composition.⁵ The use of DMS could be a more feasible and advantageous tool for routine clinical examinations due to its portability compared to large and immobile imaging methods such as dual-energy X-ray absorptiometry (DXA), computed tomography (CT), and magnetic resonance imaging (MRI). DMS provides a nonionizing imaging method that is performed at levels that do not cause increased risk of bioeffects.⁵ This approach uses DMS as a tool for assessing abdominal adiposity to demonstrate the utility of this method in quantifying an individual’s risk for metabolic conditions. The current study is a continuation of previous research that used this convenience sample of cancer survivors. Baseline measures were analyzed to assess for statistically significant associations between imaging and anthropometric measures and compare intrarater and interrater reliability of DMS measures.⁶ The current study utilizes the methodology used previously to longitudinally assess changes in body composition over time using sonographic measures and comparing them to current “gold standards” (body mass index [BMI] and DXA). By investigating the accuracy, reproducibility, and efficiency of sonographic adipose measures, this study constructs the framework for an alternative approach to help determine an individual’s risk for metabolic conditions.

Method

Eighty obese cancer survivors (4 men and 76 women) were recruited to participate in the study. All participants were known cancer survivors who had not received any active treatment for the preceding two years and met the criteria for overweight or obesity, based on BMI measurements (see Table 1). These participants were a part of a program designed to increase healthy lifestyle choices with the potential outcome of weight loss. Participants were asked to report to a sponsored clinical research center for measurements at baseline (month 0), 6 months following baseline measures (month 6), and 12 months following baseline measures (month 12). During each visit, participants underwent several, identical, standard body composition measurements. In addition, each participant had DMS and DXA imaging at each visit.

Table 1.

Body Fat Classifications Based on Body Mass Index (BMI).

Classification	BMI (kg/m²)
Healthy weight	18.5–24.9
Overweight	25–29.9
Obesity I	30–34.9
Obesity II	35–39.9
Obesity III	40 or more

Anthropometric Data

Height and weight were measured and used to calculate BMI by the following formula¹:

Weight (kg) {/ Height}^{2} (m^{2})

See Table 1. Waist circumference was measured at the superior border of the iliac crest. This measurement was conducted three times and averaged by the same registered dietitian across all three visits. DMS and DXA measures were performed concurrently at each visit.

DMS was performed by a trained sonographer using a GE logiq 9 (GE, Milwaukee, Wisconsin) ultrasound machine with a 9.0-MHz (6–8 MHz) linear transducer for participant baseline measurements. For visits 2 and 3, DMS was performed using a GE logiq i ultrasound machine with a 9.0-MHz (6–8 MHz) linear transducer. No differences in measurements were found across equipment. The protocol for scanning visceral and subcutaneous fat was obtained by using methods outlined in previous literature by Stigall et al.⁶ and Woldemariam et al.⁷ These methods included DMS postexamination measurements taken at the xiphoid, umbilicus, and the waist. Subcutaneous minimum fat and visceral fat maximum were measured at the level of the xiphoid. Subcutaneous fat maximum and visceral fat minimum were measured at the level of the umbilicus. Mesenteric fat thickness was measured at the level of the waist (Figure 1). All DMS image collection and analyses were completed by a credentialed sonographer.

Figure 1.

Diagram demonstrating the location of where sonographic measures are taken.

The DXA body composition protocol used was adopted by using methods outlined in previous literature by Stigall et al.⁶ Percent body fat (% BF) was quantified using a GE Lunar iDXA. All images and equipment were operated by a trained registered nurse. The android subscore of percent body fat (% BF), which consists of the area between the ribs and the pelvis,⁸ was chosen as the appropriate metric for comparison to DMS as it fully encompasses the abdominal region.

Data Analysis

Statistical analyses were performed by a biostatistician using SAS (version 9.4; SAS Institute, Cary, North Carolina). The associations of variables were determined by using trend graphs and Pearson correlation coefficients. The statistical power was set a priori at P < .05. Each participant’s body composition measurements were analyzed longitudinally (see Table 2 for variables). Intergroup comparisons were determined, using least squares (LS) mean differences. Changes over time in each variable were investigated using type III tests of fixed effects.

Table 2.

Descriptive Statistics for Each Measure Across Three Time Points.

Characteristic	Visit 1 (n = 67)		Visit 2 (n = 60)		Visit 3 (n = 47)
Characteristic	Mean ± SD	Range	Mean ± SD	Range	Mean ± SD	Range
Age	49.76 ± 11.42	29–72	51.12 ± 11.08	30–68	52.47 ± 10.83	30–68
Weight (kg)	92.05 ± 16.95	61.60–170.85	86.73 ± 12.78	57.80–128.70	89.29 ± 13.60	64.50–131.10
Height (m)	1.66 ± .07	1.50–1.90	1.65 ± .063	1.50–1.83	1.65 ± .063	1.50–1.83
BMI (kg/m²)	33.44 ± 5.44	25.50–61.60	31.94 ± 4.50	22.30–45.10	32.12 ± 4.45	24.9–45.9
DXA (%BF android fat)	51.75 ± 5.72	39.50–63.80	58.6 ± 20.42	38.80–126.30	50.97 ± 5.85	38.90–62.40
Waist circumference (mm)	108 ± 12.4	87.60–158.40	106 ± 11.81	77.60–138.20	105.70 ± 11.13	83.20–136.50
DMS measures (mm)
Subcutaneous minimum	1.98 ± 0.82	0.35–4.35	1.73 ± 0.60	0.83–3.63	1.59 ± 0.50	0.71–3.11
Subcutaneous maximum	3.37 ± 1.08	1.19–7.17	2.95 ± 0.99	1.21–5.26	2.91 ± 0.87	1.25–4.92
Visceral minimum	0.53 ± 0.27	0.17–1.84	0.48 ± 0.26	0.13–1.76	0.36± 0.23	0.10–0.99
Visceral maximum	1.87 ± 0.57	0.61–3.66	1.64 ± 0.49	0.58–2.79	1.49 ± 0.50	0.49–2.58
Mesenteric	4.07 ± 1.25	1.60–8.66	3.85 ± 1.05	1.47–6.56	3.61 ± 0.95	2.05–5.98

Abbreviations: BF, body fat; BMI, body mass index; DMS, diagnostic medical sonography; DXA, dual-energy X-ray absorptiometry.

Results

Due to scheduling conflicts, 77 of the 80 participants had sonographic images of their abdomen taken during at least one of their three visits. Of those participants, 67 received baseline images. In addition, 53 participants had sonographic images performed at all three visits. However, only 47 of the 53 participants’ data were used for visit 3, due to participants failing to continue to meet the inclusion criteria. Dropouts did occur across all three visits due to cancer remission and severe illness. Descriptive statistics were conducted to compare measures across all three visits. See Table 2.

The trend graphs for waist circumference, BMI, and DMS measures of subcutaneous fat demonstrated similar behavior (see Figures 2 and 3). The trend graphs for the DXA (android % BF subscore), sonographic measures of visceral fat minimum and sonographic measures of mesenteric fat demonstrated behavior similar to each other (see Figures 4 and 5).

Figure 2.

Trend graph representing body mass index (BMI) measures over time.

Figure 3.

Trend graph representing subcutaneous fat measures over time.

Figure 4.

Trend graph representing dual-energy X-ray absorptiometry (DXA) (android % body fat subscore) over time.

Figure 5.

Trend graph representing visceral fat measures over time.

Type III test of fixed effects showed significant change over time for BMI, sonographic measures of subcutaneous fat, sonographic measures of visceral fat, and sonographic measures of mesenteric fat (see Table 3). Waist circumference measures and the DXA (android % BF subscore) demonstrated no significant change over time (see Table 3). Pearson’s correlation coefficients showed significance (P < .0001–.1028) and mostly moderate to high correlations (r value range: 0.1889–0.7024) between the DXA (android % BF subscore) and visceral fat measures, subcutaneous fat measures, and mesenteric fat thickness measures for all three visits (see Table 4). In addition, BMI and waist circumference (WC) showed high correlations to subcutaneous fat and mesenteric fat thickness measures (see Table 4). Visceral minimum measures did not show a high correlation to the DXA (android % BF subscore).

Table 3.

Type III Test of Fixed Effects Where Effect Is Time.^a

Characteristic	Pr > F
BMI	<.0001
DXA (android % body fat, subscore)	.2442
Waist circumference	.3501
Subcutaneous minimum	.0007
Subcutaneous maximum	.0005
Visceral minimum	.0012
Visceral maximum	.0001
Mesenteric	.0022

Abbreviations: BMI, body mass index; DXA, dual-energy X-ray absorptiometry.

P < .05 is considered significant.

Table 4.

Pearson Correlations Comparing Visceral, Subcutaneous, and Mesenteric Fat Measures to DXA (Android % BF Subscore), BMI, and WC, With r and P Values.^a

Characteristic	P and r Values	Visceral Minimum	Visceral Maximum	Subcutaneous Minimum	Subcutaneous Maximum	Mesenteric Fat Thickness
Visit 1
DXA (android % BF subscore)	P r	.1028.2010	.0262.2716	<.0001.4582	<.0001.5022	<.0001.4959
BMI	P r	.8270.0272	.8335.0261	<.0001.5664	<.0001.6258	<.0001.6059
WC	P r	.1257.1889	.1669.1709	<.0001.4582	<.0001.5022	<.0001.4959
Visit 2
DXA (android % BF subscore)	P r	.0102.3293	.0022.3877	<.0001.4949	<.0001.7024	<.0001.5469
BMI	P r	.1303.1975	.0916.2198	<.0001.6002	<.0001.6485	<.0001.5686
WC	P r	.0213.2969	.0830.2256	<.0001.5222	<.0001.5785	<.0001.5155
Visit 3
DXA (android % BF subscore)	P r	.1164.2374	.001.4735	.0015.4591	.0002.5289	<.0001.5514
BMI	P r	.1022.2440	.0054.4035	<.0001.5888	<.0001.5685	<.0001.5459
WC	P r	.0683.2712	.0007.4804	.0002.5237	<.0001.5657	<.0001.5786

Abbreviations: BF, body fat; BMI, body mass index; DXA, dual-energy X-ray absorptiometry; WC, waist circumference.

P < .05 is considered significant.

Discussion

Assessment of abdominal fat accumulation has been investigated previously by researchers, and the most common and simplest measures used consist of BMI and WC measures.^6,8,9,10 These methods provide a noninvasive technique to assess abdominal adiposity.¹⁰ However, neither BMI nor WC can quantify variations in visceral or subcutaneous fat and therefore create pitfalls in their practical application. Unfortunately, BMI assesses the total body fat distribution and therefore does not have the ability to estimate specific abdominal adiposity measures. In addition, although insulin resistance and metabolic conditions have been associated with obesity or higher BMI, even lean subjects can be at risk for these conditions due to increased visceral fat accumulation.¹¹ The intra-abdominal fat measures by DXA and anthropometric measures compared to the visceral adipose tissues measures of single-slice CT have been explored previously by researchers.¹⁰ In a cohort of obese women, DXA and anthropometric data were inaccurate compared to CT.¹⁰ DXA has the ability to identify whole-body fat composition, similar to BMI, but lacks a way to distinguish and quantify visceral adiposity.¹⁰ Although DMS is not currently a method commonly used for quantifying abdominal adipose tissue, this study suggests that DMS may be a useful tool for assessing abdominal adipose tissue in overweight and obese cancer survivors. In this group of participants, the data suggest that all of the measures of BMI as well as the sonographic measures are more sensitive to detecting changes in body composition over time than DXA (android % BF subscore) or the WC measures.

DMS provides comparable measures of abdominal adipose tissues without the use of ionizing radiation. Arguably, obesity must be kept “in check,” maintained, and prevented through weight monitoring and weight loss therapy.¹² Mohan and Anburajan¹² conducted a study to justify the use of an anthropometric empirical indicator (AEI) by chest and pelvic radiographs for quantifying obesity. These researchers found that the use of AEI by pelvic and chest radiographs could precisely and accurately gauge obesity. The authors suggest that this method could replace the expensive and more commonly used method of DXA. Interestingly, in another study, Mauad et al.¹³ tested the accuracy and reproducibility of measuring abdominal adiposity by DMS and CT and identified an increased intraobserver reliability in DMS. However, both DMS and CT showed high accuracy and good reproducibility when measuring abdominal adiposity.¹³ Although these researchers found an alternative to DXA by using CT and AEI X-ray, these methods still required low doses of ionizing radiation, whereas DMS demonstrated high accuracy and reproducibility without exposing patients to ionizing radiation.

It is important to note that sonographic measures perform in a similar manner to the “gold standard” of DXA and BMI in determining an individual’s risk for disease. Stolk¹⁴ suggests that DMS measures performed on intra-abdominal fat of 600 participants are more reliable than WC measures. Participants’ high abdominal fat measures correlated with higher plasma glucose, total cholesterol, and triglycerides.¹⁴ The associations observed were independent of weight and contrasted with WC, suggesting that DMS measures correlate more closely with the association of abdominal adiposity and increase the risk for metabolic disease than the traditional anthropometric measures.¹⁴ Similarly, in the current study, it was shown that DMS measures were able to more accurately detect changes over time versus WC measures and android % BF subscore of DXA measures (see Table 3). In addition, WC measures correlated higher with subcutaneous fat measures than visceral fat measures. Since subcutaneous fat constitutes the majority of fat distribution within the abdomen, a high correlation with WC measures is suspected. However, as mentioned previously, WC measures cannot differentiate between different fat layers. Overestimation of subcutaneous fat or underestimation of visceral fat is possible, therefore putting patients in jeopardy of incorrectly assessing their increased risk for disease.

DMS provides a tool that can measure visceral fat accumulation that BMI and WC are unable to differentiate from total measures. Interestingly, in a study of 74 adults, DMS was found to be a valid method for assessing abdominal adiposity when measuring visceral and subcutaneous fat with strong correlations to MRI.¹⁵ Furthermore, it was found that DMS measures not only significantly improved predictions of visceral and subcutaneous fat over traditional anthropometric measures, such as BMI and WC measures, but also showed significantly stronger correlations with the MRI visceral/subcutaneous ratio, the index of abdominal adiposity used to predict metabolic conditions, over anthropometric measures.¹⁵

In the current study, BMI was highly correlated with subcutaneous fat during all three visits. This is likely due to the increased amount of subcutaneous fat distribution in individuals. The subcutaneous fat measurements showed higher correlations to the DXA (android % BF subscore) over the correlations of visceral fat. This is likely due to the limitations of measuring the visceral fat minimum. This low correlation could be due to the amount of visceral fat at the umbilicus region. This measurement is the smallest measurement of all the abdominal adiposity measurements ranging from 0.10 to 1.84 cm. Consequently, this measurement can be challenging to correlate with the DXA (android % BF subscore) since DXA measures consider the entire abdominal adiposity and are unable to distinguish between the different fat layers. Most participants demonstrated larger measures of subcutaneous fat versus visceral fat overall. Due to this increase in subcutaneous fat, it is suspected that the android % BF subscore of DXA would correlate higher with this variable, since DXA calculates the total fat distribution within the abdomen, rather than each individual layer of fat. Since the differences in fat accumulation in visceral versus subcutaneous have different health consequences, it is critical to be able to discern the two layers.

Future Implications

This longitudinal data provides an understanding that small changes can be detected in body composition over time with DMS. Therefore, the use of this data could be useful in the implementation of DMS as a screening tool for metabolic disease risk. It is important to note that this particular cohort was a convenience sample that consisted of overweight and obese cancer survivors. Cancer survivors are more susceptible to metabolic conditions and therefore may demonstrate quicker changes in adiposity measures when given an intervention. However, further research in other populations with more consistent participation is warranted. The use of DMS as a screening tool could help to determine the risk for metabolic conditions, ascertain disease recurrence, and prevent the use of added radiation to the patient. This study also provides evidence for a potential method to be used to screen pediatric abdominal adiposity in the future.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

References

Centers for Disease Control and Prevention: Defining adult overweight and obesity. 2016. www.cdc.gov/obesity/adult/defining.html. Accessed October 21, 2018.

Beilby

: Definition of metabolic syndrome: report of the National Heart, Lung, and Blood Institute/American Heart Association Conference on Scientific Issues Related to Definition. Clin Biochem Rev 2004;25(3):195–198.

Arterburn

Macjejewski

Tsevat

: Impact of morbid obesity on medical expenditures in adults. Int J Obes 2005;29:334–339.

Moore

Chaudhary

Akinyemiju

: Metabolic syndrome prevalence by race/ethnicity and sex in the United States, National Health and Nutrition Examination Survey, 1988–2012. Prev Chronic Dis 2017;14:E24.

Wagner

: Ultrasound as a tool to assess body fat. J Obes 2013;2013:280713.

Stigall

Evans

Tatarski

, et al: Abdominal adiposity measured by sonography as a tool for determining disease risk. J Diagn Med Sonography 2018;34(4):253–259.

Woldemariam

Evans

Butwin

, et al: Measuring abdominal visceral fat thickness with sonography: a methodologic approach. J Diagn Med Sonography 2018;34(2):91–96.

Stults-Kolehmainen

Stanforth

Bartholomew

Abolt

Sinha

: DXA estimates of fat in abdominal, trunk and hip regions varies by ethnicity in men. Nutr Diabetes 2013;3(3):e64.

Hamagawa

Matsumura

Kubo

, et al: Abdominal visceral fat thickness measured by ultrasonography predicts the presence and severity of coronary artery disease. Ultrasound Med Biol 2010;36(11):1769–1775.

10.

Shuster

Patlas

Pinthus

Mourtzakis

: The clinical importance of visceral adiposity: a critical review of methods for visceral adipose tissue analysis. Br J Radiol 2012;85(1009):1–10.

11.

Kahn

Prigeon

Schwartz

, et al: Obesity, body fat distribution, insulin sensitivity and islet beta-cell function as explanations for metabolic diversity. J Nutr 2001;131(2):354S–360S.

12.

Mohan

KBK

Anburajan

: Multiparametric body composition analysis and anthropometric empirical indicator. Endocrinol Invest 2013;36:672.

13.

Mauad

Chagas-Neto

Benedeti

ACGS

, et al: Reproducibility of abdominal fat assessment by ultrasound and computed tomography. Radiol Brasil 2017;50(3):141–147.

14.

Stolk

: Ultrasound measurements of intraabdominal fat estimate the metabolic syndrome better than do measurements of waist circumference. Am J Clin Nutr 2003;77(4):857–860.

15.

Rolfe

Sleigh

Finucane

, et al: Ultrasound measurements of visceral and subcutaneous abdominal thickness to predict abdominal adiposity among older men and women. Obesity 2010;18:625–631.