Does Visceral Fat Estimated by Dual-Energy X-ray Absorptiometry Independently Predict Cardiometabolic Risks in Adults?

Abstract

Background:

Abdominal visceral fat, typically measured by computer tomography (CT) or magnetic resonance imaging (MRI), has been shown to correlate with cardiometabolic risks. The purpose of this study was to examine whether a newly developed and validated visceral fat measurement from dual-energy X-ray absorptiometry (DXA) provides added predictive value to the cross-sectional differences of cardiometabolic parameters beyond the traditional anthropometric and DXA adiposity parameters.

Method:

A heterogeneous cohort of 194 adults (81 males and 113 females) with a BMI of 19 to 54 kg/m² participated in this cross-sectional study. Body composition was measured with a DXA densitometer. Visceral fat was then computed with a proprietary algorithm. Insulin sensitivity index (SI, measured by intravenous glucose tolerance test), blood pressures, and lipid profiles, and peak oxygen uptake were also measured as cardiometabolic risk parameters.

Results:

DXA-estimated visceral fat mass was associated with HDL cholesterol (regression coefficient [β] = −5.15, P < .01, adjusted R² = .21), triglyceride (β = 26.01, P < .01, adjusted R² = .14), and peak oxygen uptake (β = −3.15, P < .01, adjusted R² = .57) after adjusting for age, gender, and ethnicity. A subanalysis stratifying gender-specific BMI tertiles showed visceral fat, together with ethnicity, was independently associated with SI in overweight men and moderately obese women (second tertile).

Conclusions:

Without requiring additional CT or MRI-based measurements, visceral fat detected by DXA might offer certain advantages over the traditional DXA adiposity parameters as means of assessing cardiometabolic risks.

Keywords

cardiometabolic risks dual-energy X-ray absorptiometry insulin sensitivity visceral fat

Central obesity is associated with a number of chronic diseases, including hypertension,^1-3 dyslipidemia,^4,5 type-2 diabetes,^6,7 and is a powerful predictor for the development of cardiovascular disease^8,9 and all-cause mortality.^10,11 Abdominal subcutaneous and visceral adipose tissue are thought to differentially contribute to disease risk and all-cause mortality, with excess visceral fat having a more deleterious influence.¹⁰ Central obesity is often quantified using waist circumference, an indiscriminate measure of both subcutaneous and visceral adipose tissue that is easy to obtain. However, it can be confounded by varying levels of subcutaneous fat in the waist, and may not accurately reflect visceral fat in all individuals. Thus, accurate and feasible measurements of visceral fat are important from both public health and clinical perspectives.

Accurate measurements of visceral fat are challenging. Computed tomography (CT) has been considered the gold standard for measuring visceral fat.¹² CT generates a series of high resolution cross-sectional X-ray images through which visceral fat is identified. However, as a medical diagnostic instrument, CT scans have considerable radiation exposure. Alternatively, magnetic resonance imaging (MRI) can also measure visceral fat and has no radiation. However, both methods are still relatively expensive and limited accessibility confines their usefulness and widespread application.

Dual-energy X-ray absorptiometry (DXA) has been used to measure whole-body composition and fat distribution in heterogeneous populations over several decades. It has been explored to estimate abdominal visceral fat since the mid-90s.^13
-21 Most of these studies have established equations to estimate visceral fat area or volume with DXA-derived body composition variables (ie, total fat, android fat, etc) combined with selective anthropometry parameters. Recently, an automated image-processing algorithm was developed to estimate volumetric visceral fat using the raw 2D images from the DXA scans performed on the GE lunar iDXA.²¹ Visceral fat estimated from this algorithm has been correlated with visceral fat volume measured by multislice CT (r = .98) and is now incorporated into the GE CoreScan software. The iDXA with CoreScan is far less expensive and easier to access than CT or MRI and has less radiation exposure than CT, making it a potentially practical and valuable tool to assess visceral fat mass.

A recent cross-sectional study revealed that visceral fat measured using the iDXA and CoreScan algorithm is significantly associated with features of metabolic syndrome (blood pressure, triglyceride, high-density lipoprotein [HDL] cholesterol and fasting glucose) after adjusting for age and waist circumference.²² Other studies demonstrated that significant associations between cardiometabolic risk factors (HDL cholesterol and insulin resistance) and the visceral fat measured by the CoreScan algorithm in lean healthy Caucasian women using a GE lunar prodigy scanner,²³ and obese Caucasian and African American women using an iDXA scanner.²⁴ However, to our knowledge, visceral fat extracted from this algorithm has not been demonstrated to better predict cardiometabolic risk factors than other DXA-derived measures such as total body fat mass and android fat mass in men and women of varying body size. The purpose of this study was to examine whether a newly developed and validated visceral fat measurement from DXA provides added predictive value to interindividual differences of cardiometabolic parameters and insulin sensitivity beyond the traditional anthropometric and DXA adiposity parameters.

Methods

Subjects

A heterogeneous cross-sectional sample of 194 adults (81 males and 113 females) was recruited from a wide range of body mass index (BMI, 19-52 kg/m²) and age (42.1 ± 13.1 years) to participate in obesity-related clinical research protocols (ClinicalTrials.gov identifier: NCT00428987) at the National Institutes of Health. Subjects were admitted to the Metabolic Clinical Research Unit of the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) in Bethesda, Maryland, for a battery of clinical metabolic assessments including a DXA scan. Subjects were weight stable and had no significant diseases at the time of measurement. No female subjects were pregnant as verified by urine or serum pregnancy tests obtained less than 48 hours prior to study. Further details of the study subjects are presented in Table 1. The protocol was approved by the institutional review board of the NIDDK. Written informed consent was obtained from all subjects prior to all procedures.

Table 1.

Basic Characteristics of Study Subjects (N = 194).

	Male (n = 81)			Female (n = 113)
	Mean ± SD	Range		Mean ± SD	Range
Basic characteristics
Age, years	41.8 ± 14.1	18	67	42.4 ± 12.3	19	69
Ethnicity (African American), n (%)	33 (40.7)			55 (48.7)
Height, cm	177.2 ± 6.6	158	194	164.0 ± 7.1	143	180
Weight, kg	91.8 ± 20.9	54	152	84.4 ± 19.8	44	136
BMI, kg/m²	29.2 ± 6.5	19	45	31.4 ± 7.2	19	52
Waist circumference, cm	101.5 ± 17.4	70	139	102.7 ± 15.5	73	132
Medications
Antihypertensive, n (%)	16 (19.8)			20 (17.7)
Diabetes, n (%)	3 (3.7)			7 (6.2)
Lipid lowering, n (%)	14 (17.3)			16 (14.2)
Body composition
Total body fat, kg	29.0 ± 15.0	6.4	67.0	36.3 ± 13.6	9.5	74.1
Total body fat, %	29.9 ± 9.5	9.9	50.9	41.6 ± 7.9	16.3	54.7
Android fat, kg	2.96 ± 1.93	0.23	7.77	3.15 ± 1.62	0.34	7.64
Android fat, %	36.9 ± 14.5	7.3	62.6	46.1 ± 12.0	9.8	65.1
Visceral fat, kg	1.44 ± 1.17	0.07	4.89	0.95 ± 0.68	0.05	3.11
Cardiometabolic risks
Insulin sensitivity index	5.3 ± 4.2	0.17	18.4	4.4 ± 3.1	0.46	16.7
Systolic blood pressure, mmHg	126.8 ± 12.2	100	150	126.6 ± 12.5	95	160
Diastolic blood pressure, mmHg	75.9 ± 9.0	60	98	72.7 ± 9.1	50	96
Total cholesterol, mg/dL	169.9 ± 36.6	91	255	181.6 ± 38.1	114	296
HDL cholesterol, mg/dL	46.0 ± 13.8	26	107	52.9 ± 16.2	21	100
LDL cholesterol, mg/dL	98.4 ± 29.0	36	167	106.4 ± 32.7	51	213
Triglycerides, mg/dL	127.1 ± 93.0	39	635	113.7 ± 66.9	24	392
Peak oxygen uptake, ml/kg/min	30.5 ± 11.1	12	63	22.6 ± 7.1	10	42

BMI, body mass index; HDL, high-density lipoprotein; LDL, low-density lipoprotein.

Demographic and Anthropometric Measurements

Age, gender, ethnicity (self-reported), BMI and waist circumference were recorded. Body weight was measured using a digital balance (Scale-Tronix 5702, Carol Stream, IL). Height was measured using a stadiometer (Seca 242, Hanover, MD). BMI was calculated as body weight in kilograms divided by the squared height in meters. Waist circumference was measured to the nearest 0.1 cm at the midpoint between the lowest rib and the iliac crest. All anthropometrics were collected by trained study staff.

DXA Measurements

Body composition was measured using a total body scanner (Lunar iDXA, GE Healthcare, Madison WI). The iDXA is a narrow-fan beam instrument with a high weight limit (200 kg limit) and a relatively wide field of view (66 cm) designed to accommodate obese subjects. Traditional body composition analysis was performed using GE enCore 11.10 software. Total (regional) fat percentage was computed as follows: (total fat mass / (fat mass + lean soft mass + bone mineral content)) × 100. For measuring android fat, a region-of-interest is automatically defined as between the top of the iliac crest and 20% of the distance from the top of the iliac crest to the base of the skull. Android fat percentage was calculated as follows: (android fat mass / (android fat mass + android lean mass + android bone mineral content)) × 100. Precisions of iDXA for total and android fat masses are reported to be coefficients of variations (CV) of 0.8-1.0% and 2.3-2.4%, respectively.^25,26

Estimation of Visceral Fat

The approach of the proprietary visceral fat algorithm was previously described.^21,27 Briefly, the android region contains both visceral and subcutaneous fats. The subcutaneous fat forms a layer of nonuniform thickness around the abdominal cavity. X-ray attenuation in the DXA soft-tissue image is used to determine the edge of the body and the outer edge of the abdominal cavity. The distance between the 2 edges is used to determine the width of the subcutaneous fat layer along the lateral extent of the abdomen. The anteroposterior abdominal thickness of the abdomen is derived from basis set transformation using the dual-energy attenuation spectrum information.²⁷ The subcutaneous fat width and anteroposterior abdominal thickness along with empirically derived geometric constants are then used to calculate the quantity of subcutaneous fat in the android region. Visceral fat is computed by subtracting subcutaneous fat from the total android fat mass in the android region. The algorithm is now incorporated into the GE CoreScan software option.

Operationally, we anonymized the raw scans from our DXA database (2007-2011) with only linked information, and sent the dataset to GE Global Research Center where the visceral fat was analyzed and returned. We further limited subjects to those with calculated visceral fat greater than 42 cm³, which was thought to be the lower limit for detecting visceral fat according to the original development paper.²¹ A recent study showed that the precision in a clinical sample of 32 obese women who replicated scans with repositioning was CV of 5.1%.²⁸ As an internal validation of this study, we also tested the reproducibility of this algorithm in 68 subjects (24 male and 44 female) who had undergone a repeated DXA scan within 14 days and found an intraclass correlation and coefficient of variation to be 0.99 and 9.8% respectively (see Supplemental Material 1 available at dst.sagepub.com/supplemental).

Cardiometabolic Measures

Insulin sensitivity index (SI, measured by intravenous glucose tolerance test [IVGTT]), blood pressures (standard mercury sphygmomanometer), total, HDL and low-density lipoprotein (LDL) cholesterols, triglycerides and peak oxygen uptake (peak VO₂) were collected as presented below.

Insulin sensitivity index

The IVGTT was performed in the morning after a 12-hour overnight fast with 2 intravenous catheters placed in each antecubital vein. At time 0, an intravenous bolus of glucose (0.3 mg/kg), was administered over 1 minute. A bolus of insulin (0.03 units/kg for nondiabetics, 0.05 for diabetic patients) was given at 20 minutes. Blood samples (3.0 cc) for glucose, insulin, and free fatty acids were taken at −10, −1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 10, 12, 14, 16, 20, 22, 23, 24, 25, 27, 30, 40, 50, 60, 70, 80, 90, 100, 120, 150, and 180 minutes. SI was calculated using the MinMod Millennium (6.02) program.²⁹

Blood pressures

Under a 12-hour overnight fasted condition, trained nurses measured systolic and diastolic blood pressures using a standard mercury sphygmomanometer on the right arm of seated participants who had rested for at least 10 minutes.

Blood lipids

A blood sample was drawn from each participant after an overnight (≥ 12 h) fast. Routine assays for serum lipids were performed in the Department of Laboratory Medicine at the Clinical Center, NIH. Serum total cholesterol and triglycerides were determined enzymatically, and serum HDL cholesterol was measured by the heparin-manganese precipitation method. Serum LDL cholesterol was estimated according to the method of Friedewald et al.³⁰

Peak oxygen uptake

Peak VO₂ was determined by a graded exercise test using a cycling ergometer. Following a 2-min warm up at 0 watts, the workload ramped up by 10, 15, 20, or 25 watts per minute dependent on subject’s gender, size, and self-reported fitness level until volitional exhaustion. The tests were typically completed within 12 minutes. During the test, ventilation and expiratory gases were measured using an indirect calorimeter (TrueOne 2400, Parvomedics, Sandy, UT). The highest oxygen uptake achieved over 30 seconds was determined as the peak VO₂.

Data and Statistical Analysis

Summary data for all continuous variables were shown in mean ± standard deviations. To explore contribution of DXA-derived visceral fat to cardiometabolic risk factors, we applied multiple regression analyses with backward elimination where Akaike’s information criterion (AIC) was chosen as selection criterion. In these analyses, we entered each cardiometabolic risk factor as the dependent variable with age, gender, ethnicity as fixed covariates and visceral fat, android fat mass, total fat mass, BMI, and waist circumference as independent candidate variables in the model.

A second, exploratory analysis was performed to investigate the idea that visceral fat may be more sensitive to predict cardiometabolic risks in overweight and moderately obese individuals than in very lean and/or severely obese subjects. To accomplish this, we stratified our heterogeneous sample by BMI, dividing our data into 3 groups based on the gender-specific BMI tertiles and combined each of the same tertile from both genders (ie, male first tertile combined with female first tertile) to maintain robust statistical power. We also applied multiple regression approaches as stated above for this subanalysis.

In the regression analyses, SI was log-transformed to achieve a normal distribution. Subjects who had no measurement, missing values, or had medical conditions or were taking prescription medication that could affect the outcome variables were eliminated from each analysis on a per-outcome basis. Specifically, out of the 194 adults included in this study, 10 diabetic patients and 43 subjects who had did not undergo the IVGTT were excluded from analyses with SI, resulting in 141 subjects’ data ultimately used in these analyses. Similarly, 145 for systolic blood pressure, 144 for diastolic blood pressure, 130 for total cholesterol, 133 for HDL cholesterol, 128 for LDL cholesterol, 129 for triglycerides, and 187 adults for peak VO₂ were finally included in each analysis as described in Tables 2 and 3.

Table 2.

Contribution of DXA-Derived Visceral Fat to Cardiometabolic Risk Factors.

	Adj. R²	BMI		Waist		Total fat		Android fat		Visceral fat
	Adj. R²	β	P	β	P	β	P	β	P	β	P
Log insulin sensitivity index (n = 141)	0.54	—	—	—	—	—	—	—0.32	<0.01	—	—
Systolic blood pressure (n = 145)	0.14	—	—	0.29	<0.01	—	—	—	—	—	—
Diastolic blood pressure (n = 144)	0.08	—	—	–0.26	0.09	—	—	3.36	0.01	—	—
Total cholesterol (n = 130)	0.33	—	—	0.57	<0.01	—	—	—	—	—	—
HDL cholesterol (n = 133)	0.21	—	—	—	—	—	—	—	—	–5.15	<0.01
LDL cholesterol (n = 128)	0.22	—	—	0.53	<0.01	—	—	—	—	—	—
Triglycerides (n = 129)	0.14	—	—	—	—	—	—	—	—	26.01	<0.01
Peak oxygen uptake (n = 187)	0.57	0.47	.02	—	—	–0.39	<0.01	—	—	–3.15	<0.01

Adj. R², adjusted R² value; β, regression coefficient; HDL, high-density lipoprotein; LDL, low-density lipoprotein; —, not selected in the final models.

Table 3.

Contribution of DXA-Derived Visceral Fat to Insulin Sensitivity Index According to Gender-Specific Tertiles (Model Selection With Backward Elimination Using 3 Fixed Covariates and 5 Independent Variables).

	Tertile 1 (n = 47), Adj. R² = 0.28		Tertile 2 (n = 47), Adj. R² = 0.43		Tertile 3 (n = 47), Adj. R² = 0.44
	β	P	β	P	β	P
Independent variables
Visceral fat, kg	—	—	–0.42	0.03	0.29	0.10
Android fat, kg	—	—	–0.27	0.07	–0.61	<0.01
Total fat, kg	—	—	—	—	—	—
BMI, kg/m²	–0.09	<0.01	—	—	—	—
Waist, cm	—	—	—	—	—	—
Fixed covariates
Age, years	0.00	0.58	0.01	0.05	0.02	0.03
Gender (men)	0.12	0.39	0.19	0.36	–0.41	0.07
Ethnicity (African American)	–0.25	0.16	–0.60	<0.01	–0.56	<0.01

Adj. R², adjusted R² value; β, regression coefficients; —, not selected in the final model. Tertile 1: BMI of 18.9-26.3 kg/m² for men and 19.2-29.0 kg/m² for women. Tertile 2: BMI of 26.7-30.6 kg/m² for men and 29.2-35.9 kg/m² for women. Tertile 3: BMI of 30.7-45.1 kg/m² for men and 35.9-45.9 kg/m² for women.

All data handling and statistical analyses were performed with the open-source R (http://www.r-project.org/), and statistical significance was set at P < .05.

Results

Subject Characteristics

Our study population represented a heterogeneous population (descriptive statistics shown in Table 1) with a wide spectrum of age, adiposity, and cardiorespiratory fitness level. Among these subjects, 18.6%, 5.5%, and 15.5% of subjects were on treatment for hypertension, diabetes, and dyslipidemia, respectively. These subjects were excluded from specific analyses.

Contributions of DXA-Derived Visceral Fat to Cardiometabolic Risk Factors

The final models with least AIC value revealed that waist circumference was associated with systolic blood pressure, and total and LDL cholesterols after adjusting for the fixed covariates of age, gender and ethnicity (Table 2). Android fat was associated with insulin sensitivity and diastolic blood pressure whereas visceral fat was associated with HDL cholesterol, triglyceride, and peak VO₂. A 1-kg increase in visceral fat corresponds to a 5.1 mg/dL decrease in HDL cholesterol, 26.1 mg/dL increase in triglyceride, and a 3.2 ml/kg/min decrease in peak VO₂.

Subanalysis With Gender-Specific BMI Tertiles

When we further explored the sample by stratifying into gender-specific BMI tertiles, DXA-derived visceral fat, together with ethnicity, were independently associated with SI in tertile 2 (BMI 26.7-30.6 kg/m² for men and 29.2-35.9 kg/m² for women). In contrast, in tertile 1, visceral fat did not play an independent role in determining SI (as shown in Table 3). In tertile 3, visceral fat was selected in the final model but was not significant in predicting SI (P = .10).

Discussion

Increased body fatness leads to increased cardiometabolic risks, including type 2 diabetes. Central adiposity, and more specifically, abdominal visceral fat, is thought to be more directly linked to these risk factors.¹⁰ While “low-tech” measurements, such as BMI and waist circumferences, are important for public health screening and large-scale surveys, accurate and precise determinations of visceral fat may offer important additional information during clinical assessments. Currently, visceral fat is traditionally quantifiable through CT or MRI scans. However, these methods are costly, may involve rigorous postprocessing, and, in the case of CT, involve substantial exposure to more than minimum radiation (around 3 mSV).²¹

Aided by a newly developed analysis tool to estimate visceral fat mass from 2-dimensional scans using the iDXA, our study examined whether the iDXA-derived visceral fat measure provides additional predictive value for cardiometabolic risk factors beyond the traditional DXA measured adiposity and other standard anthropometric measures. We found that increased visceral fat was more highly correlated with low HDL cholesterol, elevated triglyceride level and low peak VO₂ than traditional DXA measures such as total and android fat masses, and was associated with the decreased insulin sensitivity in overweight men and moderately obese women. Consequently, the substantially lower cost and radiation of the DXA scan (around 1 µSV)²¹ may be useful in the identification and management of deleterious levels of visceral adipose tissue.

Our findings that android fat mass was independently associated with some features of metabolic syndrome were consistent with several previous studies. Lee et al³¹ demonstrated the comparable contributions of DXA-derived abdominal fat and CT-derived intra-abdominal fat area on predicting metabolic syndrome in obese women. Paradisi et al³² revealed that DXA-determined adiposity in the abdominal region between L1-4 level, possibly similar to our android fat parameters (mass and percentage), was the most predictive of the metabolic variables, showing significant relationships with glucose infusion rate, triglyceride, and cholesterol, and such effects were independent of total body mass. Furthermore, android fat mass was significantly associated with clustering of metabolic syndrome components after adjustment for multiple parameters including age, gender, adiponectin, high-sensitivity C reactive protein, a surrogate marker of insulin resistance, whole-body fat mass, and even visceral fat area.³³

A recent study revealed that DXA visceral fat measured using the same software algorithm and scanner-type used in this study was associated with cardiometabolic risks in a cross-sectional population.²² However, this study controlled only for age, BMI, and waist circumference, but not the traditional DXA estimates. Our study showed that iDXA estimated visceral fat did add to DXA-derived total and android fat mass in explaining difference in cardiometabolic outcomes.

Kang et al³³ demonstrated potential fat infiltration of the liver, pancreas, and other organs by visceral fat using CT and that these organs are important in the production and processing of cholesterol, which may help to explain the association that we saw between iDXA-derived visceral fat, lower HDL cholesterol and elevated triglyceride level.

Peak VO₂ is generally considered to reflect long-term physical activity status. Previous studies revealed that physically active individuals have lower abdominal visceral fat accumulation than inactive counterparts^34
-36 and an intervention study has shown that a reduction in visceral fat due to hypocaloric diet combined with walking is significantly associated with peak VO₂ improvement.³⁷ This may help to explain the significant association between increased visceral fat and low peak VO₂.

We found that visceral fat acted as an independent predictor of insulin sensitivity only in individuals falling in the second BMI tertile for our cohort. Not surprisingly, we also observed that a substantial proportion of individuals in the lowest BMI tertile of our generally healthy cohort had very little visceral fat (<1.2 kg), suggesting a possible “floor effect” in the visceral fat and SI in this group. Conversely, the predominating low SI in the highest BMI tertile may yield a “ceiling effect” to the visceral fat. Taken together, these subgroup analyses suggested a potential nonlinear interaction between visceral fat and SI, which may not be evident when traditional linear-based analyses, such as the multiple regression analysis presented here, are applied to the entire cohort.

To demonstrate internal validity of the iDXA visceral fat measure, we tested the reproducibility of this algorithm using subsample of the clinical population using a test-retest paradigm. The correlation between 2 scans of the same individual separated by less than 2 weeks was excellent (r = .98) and CV was reasonable (9.8%) but larger than the previously reported variability of the same instrument and visceral fat algorithm (5.1%).²⁸ This difference may be attributed by a procedural difference and/or by the dynamic range in the BMI of the subjects. Rothney et al scanned subjects with the maximum BMI of 40 kg/m² twice in the same day with repositioning²⁸ while we performed scans twice on 2 separate days in subjects BMI up-to 51 kg/m². Thus, our results might include subtle physiological changes in body composition that naturally occurred even though there was no significant weight change or treatment between the 2 scan days.

This study has several strengths. First, the study population incorporated a wide range of adiposity levels of men and women which also included African Americans. Previous studies with the same algorithms included mainly a BMI range of 18-40 kg/m² of Caucasians men and women²² or in obese women only.²⁴ Second, all tests were performed in a well-controlled inpatient settings where we could carefully measure cardiometablic parameters such as insulin sensitivity and maintain hydration status for body composition analysis. We determined that the iDXA-derived visceral fat may provide additional information to explain cardiometabolic risks compared to currently available DXA estimates. Although similar findings have been shown using CT, the DXA-derived visceral fat may offer wider application due to its lower radiation exposure and quicker analysis time.

There were also several limitations to the current study. We were unable to prove causal relationships between DXA-derived visceral fat and cardiometabolic risks due to the cross-sectional study design. Longitudinal and intervention studies are necessary to explore causal links and to confirm our findings. With 194 subjects, a relative small sample size was another limitation of this study. Moreover, not all DXA scanners can be used to estimate visceral fat. To our knowledge, only scans performed by narrow-angle fan beam GE systems (Lunar Prodigy and iDXA, both from GE Healthcare, Madison, WI) with recent versions of the enCore software can be used for this specific analysis (personal communications with GE Healthcare engineering support). The CoreScan software is an optional add-on which can now be purchased with the standard enCore DXA operational software. Neither the pencil-beam systems (such as the Lunar DPX) nor their scans can be used or reanalyzed for visceral fat estimations. In addition, the visceral fat analysis requires the android region to be clearly in the field of view where the external wall cannot be in contact with the arm or hands. The Hologic Discovery DXA System (Hologic, Bedford, MA) also has a similar software package (InnerCore™) that allows for the assessment of visceral fat using a similarly described approach,^20,38 which can be applied to scans performed with all modern Hologic fan-beam technology.³⁸

Conclusions

The current study indicated that visceral fat detected by the novel algorithms with iDXA was associated with HDL cholesterol, triglyceride, and peak VO₂ beyond traditional DXA measures such as total and android fat masses. Visceral fat estimates from iDXA have some advantages over traditional DXA measures as a means of assessing caridiometabolic risks, although the magnitude of the additional contributions was fairly small and was limited to only peak VO₂, HDL cholesterol, triglyceride, and SI in overweight men and moderately obese women. For merely 0.03% of the radiation exposure of an abdominal CT, the DXA-measured visceral fat may be useful as a screening tool for subjects or patients at risk for metabolic syndrome, and may be used repeatedly in prospective studies to track changes in visceral fat.

Footnotes

Abbreviations

AIC, Akaike’s information criterion; BMI, body mass index; CT, computed tomography; CV, coefficient of variance; DXA, dual-energy X-ray absorptiometry; HDL, high-density lipoprotein; IVGTT, intravenous glucose tolerance test; LDL, low-density lipoprotein; MRI, magnetic resonance imaging; NIDDK, National Institute of Diabetes and Digestive and Kidney Diseases; NIH, National Institutes of Health; peak VO₂, maximal oxygen uptake; SI, insulin sensitivity index.

Declaration of Conflicting Interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: MPR was working for GE Global Research Center while the original analysis of visceral fat was performed.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Intramural Research Programs of NIDDK (1ZIA DK075007, 1ZIA DK071013, and 1ZIA DK071014).

References

Kanai

Matsuzawa

Kotani

. Close correlation of intra-abdominal fat accumulation to hypertension in obese women. Hypertension. 1990;16:484-490.

Koh

Hayashi

Sato

. Visceral adiposity, not abdominal subcutaneous fat area, is associated with high blood pressure in Japanese men: the Ohtori study. Hypertens Res. 2011;34:565-572.

Hayashi

Boyko

Leonetti

. Visceral adiposity and the prevalence of hypertension in Japanese Americans. Circulation. 2003;108:1718-1723.

Liu

Fox

Hickson

. Impact of abdominal visceral and subcutaneous adipose tissue on cardiometabolic risk factors: the Jackson Heart Study. J Clin Endocrinol Metab. 2010;95:5419-5426.

Fox

Massaro

Hoffmann

. Abdominal visceral and subcutaneous adipose tissue compartments: association with metabolic risk factors in the Framingham Heart Study. Circulation. 2007;116:39-48.

Hayashi

Boyko

Leonetti

. Visceral adiposity and the risk of impaired glucose tolerance: a prospective study among Japanese Americans. Diabetes Care. 2003;26:650-655.

Hayashi

Boyko

McNeely

Leonetti

Kahn

Fujimoto

. Visceral adiposity, not abdominal subcutaneous fat area, is associated with an increase in future insulin resistance in Japanese Americans. Diabetes. 2008;57:1269-1275.

Mahabadi

Massaro

Rosito

. Association of pericardial fat, intrathoracic fat, and visceral abdominal fat with cardiovascular disease burden: the Framingham Heart Study. Eur Heart J. 2009;30:850-856.

Fujimoto

Bergstrom

Boyko

. Visceral adiposity and incident coronary heart disease in Japanese-American men. The 10-year follow-up results of the Seattle Japanese-American Community Diabetes Study. Diabetes Care. 1999;22:1808-1812.

10.

Katzmarzyk

Mire

Bouchard

. Abdominal obesity and mortality: the Pennington Center Longitudinal Study. Nutr Diabetes. 2012;2:e42.

11.

McNeely

Shofer

Leonetti

Fujimoto

Boyko

. Associations among visceral fat, all-cause mortality, and obesity-related mortality in Japanese Americans. Diabetes Care. 2012;35:296-298.

12.

Rössner

Hiltbrandt

. Adipose tissue determinations in cadavers–a comparison between cross-sectional planimetry and computed tomography. Int J Obes. 1990;14:893-902.

13.

Jensen

Kanaley

Reed

Sheedy

. Measurement of abdominal and visceral fat with computed tomography and dual-energy X-ray absorptiometry. Am J Clin Nutr. 1995;61:274-278.

14.

Clasey

Bouchard

Teates

. The use of anthropometric and dual-energy X-ray absorptiometry (DXA) measures to estimate total abdominal and abdominal visceral fat in men and women. Obes Res. 1999;7:256-264.

15.

Kamel

McNeill

Van Wijk

. Usefulness of anthropometry and DXA in predicting intra-abdominal fat in obese men and women. Obes Res. 2000;8:36-42.

16.

Park

Y-W

Heymsfield

Gallagher

. Are dual-energy X-ray absorptiometry regional estimates associated with visceral adipose tissue mass? Int J Obes Relat Metab Disord. 2002;26:978-983.

17.

Stewart

DeRegis

Turner

. Usefulness of anthropometrics and dual-energy X-ray absorptiometry for estimating abdominal obesity measured by magnetic resonance imaging in older men and women. J Cardiopulm Rehabil. 2003;23:109-114.

18.

Hill

LaForgia

Coates

Buckley

Howe

PRC

. Estimating abdominal adipose tissue with DXA and anthropometry. Obesity (Silver Spring). 2007;15:504-510.

19.

Gradmark

AMI

Rydh

Renström

. Computed tomography-based validation of abdominal adiposity measurements from ultrasonography, dual-energy X-ray absorptiometry and anthropometry. Br J Nutr. 2010;104:582-588.

20.

Micklesfield

Goedecke

Punyanitya

Wilson

Kelly

. Dual-energy X-ray performs as well as clinical computed tomography for the measurement of visceral fat. Obesity (Silver Spring). 2012;20:1109-1114.

21.

Kaul

Rothney

Peters

. Dual-energy X-ray absorptiometry for quantification of visceral fat. Obesity (Silver Spring). 2012;20:1313-1318.

22.

Rothney

Catapano

Xia

. Abdominal visceral fat measurement using dual-energy X-ray: association with cardiometabolic risk factors. Obesity (Silver Spring). 2013;21:1798-1802.

23.

Miazgowski

Krzyżanowska-Świniarska

Dziwura-Ogonowska

Widecka

. The associations between cardiometabolic risk factors and visceral fat measured by a new dual-energy X-ray absorptiometry-derived method in lean healthy Caucasian women. Endocrine. 2014;47:500-505.

24.

Seabolt

Shibao

. DXA-measured visceral adipose tissue predicts impaired glucose tolerance and metabolic syndrome in obese Caucasian and African-American women. Eur J Clin Nutr. 2014;1-8.

25.

Hind

Oldroyd

Truscott

. In vivo precision of the GE Lunar iDXA densitometer for the measurement of total body composition and fat distribution in adults. Eur J Clin Nutr. 2011;65:140-142.

26.

Rothney

Martin

F-P

Xia

. Precision of GE Lunar iDXA for the measurement of total and regional body composition in nonobese adults. J Clin Densitom. 2012;15:399-404.

27.

Ergun

Rothney

Oates

Xia

Wacker

Binkley

. Visceral adipose tissue quantification using lunar prodigy. J Clin Densitom. 2013;16:75-78.

28.

Rothney

Xia

Wacker

. Precision of a new tool to measure visceral adipose tissue (VAT) using dual-energy X-Ray absorptiometry (DXA). Obesity (Silver Spring). 2013;21:E134-E136.

29.

Boston

Stefanovski

Moate

Sumner

Watanabe

Bergman

. MINMOD Millennium: a computer program to calculate glucose effectiveness and insulin sensitivity from the frequently sampled intravenous glucose tolerance test. Diabetes Technol Ther. 2003;5:1003-1015.

30.

Friedewald

Levy

Fredrickson

. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem. 1972;18:499-502.

31.

Lee

Kim

Y-J

Kim

Y-J

. Waist circumference, dual-energy X-ray absortiometrically measured abdominal adiposity, and computed tomographically derived intra-abdominal fat area on detecting metabolic risk factors in obese women. Nutrition. 2008;24:625-631.

32.

Paradisi

Smith

Burtner

. Dual energy X-ray absorptiometry assessment of fat mass distribution and its association with the insulin resistance syndrome. Diabetes Care. 1999;22:1310-1317.

33.

Kang

Yoon

Ahn

. Android fat depot is more closely associated with metabolic syndrome than abdominal visceral fat in elderly people. PLoS One. 2011;6:e27694.

34.

Kay

Fiatarone Singh

. The influence of physical activity on abdominal fat: a systematic review of the literature. Obes Rev. 2006;7:183-200.

35.

Ohkawara

Tanaka

Miyachi

Ishikawa-Takata

Tabata

. A dose-response relation between aerobic exercise and visceral fat reduction: systematic review of clinical trials. Int J Obes (Lond). 2007;31:1786-1797.

36.

Ismail

Keating

Baker

Johnson

. A systematic review and meta-analysis of the effect of aerobic vs. resistance exercise training on visceral fat. Obes Rev. 2012;13:68-91.

37.

Lynch

Nicklas

Berman

Dennis

Goldberg

. Reductions in visceral fat during weight loss and walking are associated with improvements in VO₂max. J Appl Physiol. 2001;90:99-104.

38.

Kelly

. Practical and technical advantages of DXA visceral fat assessment compared with computed tomography. Hologic Inc. Available at: http://www.hologic.com/sites/default/files/white-papers/WP_00062_Visceral_Fat_06-12.pdf. Accessed February 3, 2015.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.28 MB