Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Am J Clin Nutr. Author manuscript; available in PMC 2010 January 4.
Published in final edited form as:
PMCID: PMC2801427

Visceral adiposity and its anatomical distribution as predictors of the metabolic syndrome and cardiometabolic risk factor levels



Despite the recognition that central obesity plays a critical role in chronic disease, few large-scale imaging studies have documented human variation in abdominal adipose tissue patterning.


We aimed to compare the associations between abdominal subcutaneous adipose tissue (ASAT) and visceral abdominal tissue (VAT), which were measured at different locations across the abdomen, and the presence of the metabolic syndrome (MS; National Cholesterol Education Program Adult Treatment Panel III definition) and individual cardiometabolic risk factors.


This study included 713 non-Hispanic whites aged 18–86 y, in whom VAT and ASAT were assessed by using multiple-image magnetic resonance imaging. The anatomical position of the magnetic resonance image containing the maximum VAT area for each subject was used as a measure of VAT patterning. Multivariate linear and logistic regression analyses were used to examine the relation of VAT, ASAT, and VAT patterning to cardiometabolic risk.


VAT mass was a stronger predictor of the MS than was ASAT mass, but ASAT mass (and other measures of subcutaneous adiposity) had signification interactions with VAT mass, whereby elevated ASAT reduced the probability of MS among men with high VAT (P = 0.0008). There was variation across image locations in the association of VAT area with the MS in men, and magnetic resonance images located 4–8 cm above L4–L5 provided the strongest correlations between VAT area and cardiometabolic risk factors. Subjects whose maximum VAT area was higher in the abdomen had higher LDL-cholesterol concentrations (R2 = 0.07, P = 0.0001), independent of age and adiposity.


Further studies are needed to confirm the effects of VAT patterning on cardiometabolic risk.


Abdominal obesity predisposes persons to diabetes and cardiovascular disease (CVD) (1), and elevated visceral adipose tissue (VAT), in particular, is associated with insulin resistance, dyslipidemia, systemic inflammation, diabetes, hypertension, myocardial infarction, and all-cause mortality (214). Whereas VAT may have a uniquely important pathophysiological role in CVD and diabetes, the amount of subcutaneous adipose tissue (SAT) also contributes to obesity-related insulin resistance and the metabolic syndrome (MS), as reviewed by Misra and Vikram (7) and Freedland (15). Indeed, it has been suggested that low SAT mass in the limbs may increase the risk of diabetes (1, 1620). However, the role of abdominal SAT (ASAT) in buffering the effects of VAT on health is not entirely clear.

Because of cost constraints, most imaging studies have obtained a single magnetic resonance imaging (MRI) or computed tomography image of the abdomen, typically at the L4–L5 intervertebral space, to represent VAT and abdominal SAT (21). However, studies using multiple-image protocols suggest that, compared with images taken in the middle to upper abdomen, ie, near L2–L3, the L4–L5 image is a significantly worse predictor of both total VAT volume (2225) and CVD risk factor levels and the MS (23, 26, 27). Furthermore, individual variation in the spatial distribution of the VAT depot itself in the upper and lower abdomen has received relatively little attention. Machann et al (28) elegantly illustrated variation in adipose tissue “topography,” but epidemiologic studies have yet to examine the spatial patterning of VAT as an independent predictor of health risks.

With the use of a large database of 1 cm–thick contiguous abdominal MRI images, the present study aimed 1) to examine effects of VAT and ASAT and their interaction with each other on the presence or absence of the MS, 2) to determine whether image location significantly affects the correlations between single image VAT areas and cardiometabolic risk factor levels and the presence or absence of the MS, and 3) to explore variation in the distribution of VAT across the abdomen and its relation to cardiometabolic risk factor levels and the MS.


Study sample

The sample included 713 healthy non-Hispanic white participants (49% women) aged 18–86 y who were enrolled in 1 of 3 ongoing studies of body composition and CVD risk—the Fels Longitudinal Study, the Miami Valley Family Aging Study, and the Southwest Ohio Family Study—at the Lifespan Health Research Center (Dayton, OH) in 2003–2006. Participants resided in the Dayton-Cincinnati area of southwestern Ohio at the time of recruitment and returned to the study center at regular intervals for serial examinations. Participants were screened in advance to ensure that they were free of any contraindications for MRI. All of the data for a scheduled exam were collected on the same day.

All subjects provided written informed consent. The study protocol was approved by the Institutional Review Board at Wright State University.

Anthropometry and whole-body body composition

Participants wore light clothing (eg, shorts and sleeveless shirts) during measurement. Weight was measured to the nearest 0.01 kg and stature was measured to the nearest 0.01 cm by using a digital scale and a digital stadiometer, respectively. Total-body fat mass (TBF), fat-free mass (FFM), and lower-extremity (LEG) fat mass were measured by using a dual-energy X-ray absorptiometry (DXA) system (Hologic QDR 4500; Hologic Inc, Bedford, MA; software version 9.8D). The triceps subcutaneous skinfold thickness was measured by using Lange calipers.

Magnetic resonance imaging

Abdominal MRI was conducted at the Good Samaritan Hospital Greater Dayton MRI Consortium (Dayton, OH) by using a previously described protocol (22, 29). Images were obtained every 1 cm from the 9th thoracic vertebra (T9) to the first sacral vertebra (S1). Image locations were defined relative to a common anatomical landmark, the L4–L5 intervertebral space. To facilitate comparison of individual image data, we limited our analyses to the set of images with no missing values for all subjects, and this set included images ranging from 20 cm above L4–L5 (+20 cm) to 3 cm below L4–L5 (−3 cm), for a total of 24 images per subject. Segmentation of the axial images into VAT and ASAT areas (cm2) was performed by 2 trained observers using SLICE-O-MATIC image analysis software (version 4.2; Tomo-vision Inc, Montreal, Canada). Interobserver variation was low (CVs=5.1% for VAT mass and 1.2% for ASAT mass). VAT and SAT areas were summed across all 24 images to obtain VAT and SAT volumes, and then these volumes were multiplied by 0.916 g/cm3, the density of adipose tissue, to obtain total VAT mass (kg) and total ASAT mass (kg). The spatial distribution, or patterning, of VAT across the abdomen was assessed by identifying the maximum (peak) VAT area for each subject, and recording the measurement location of that image. Individual maximum VAT areas occurred across a broad range of images, from 3 cm below L4–L5 to 19 cm above L4–L5. On average, maximum VAT area occurred 6.7 cm above L4–L5 in men and 3.8 cm above L4–L5 in women (P for sex difference < 0.0001).

Cardiometabolic risk factor measures and the metabolic syndrome

Blood samples were obtained after a 12-h fast and collected by using a evacuated tube containing EDTA. Fasting plasma glucose concentrations were measured in-house by using the Vitros Chemistry Products VITROS GLU DT method (Ortho-Clinical Diagnostics, Raritan, NJ) on the VITROS DT60 analyzer. Fasting triglycerides (TG) and the HDL- and LDL-cholesterol fractions were measured at the Medical Research Laboratory (Cincinnati, OH), which is participating in the Centers for Disease Control and Prevention's Lipid Standardization Program. Fasting insulin, measured by radioimmunoassay, was available for a subset of subjects. The homeostasis model assessment score, an estimate of insulin resistance, was calculated from fasting insulin and glucose values (30). High-sensitivity C-reactive protein was measured in a subset of participants with the use of an immunochemiluminometric method. Trained technicians measured waist circumference to the nearest centimeter immediately superior to the left iliac crest according to protocols of the National Health and Nutrition Examination Survey. Seated blood pressures were measured with a mercury sphygmomanometer after 5 min of rest, and appropriate cuff size was chosen on the basis of the arm circumference of the participant. The means of the last 2 of 3 seated systolic and diastolic blood pressure measurements were used for analysis. The MS was defined according to the updated criteria of the National Cholesterol Education Program Adult Treatment Panel III (31). The prevalence of the MS was significantly higher in men than women (31.2% and 24.0%, respectively; P = 0.02). Sex differences also were found in the prevalence of all MS components (P<0.05, data not shown) and in all body-composition measures, and therefore analyses were stratified by sex.

Statistical analysis

Skewness and kurtosis of the study variables were examined, and variables were transformed by using the natural log function only when necessary to meet statistical assumptions for general linear models (32). The sample initially included 776 subjects who underwent abdominal MRI; 22 were removed because of diabetes (11 men and 11 women), and 41 subjects with values >3 SD above the mean for >1 cardiometabolic risk factor were removed (25 men and 16 women), to yield the final sample of 713. Differences in means between the sexes were tested by using unpaired t tests.

Sex-specific logistic regression models adjusted for age, age2, body mass index (BMI; in kg/m2), current smoking status (yes or no), physical activity level [using the sport index of the Baecke questionnaire (33)], and FFM were tested to examine the effect of VAT, ASAT, and BMI on the odds of the MS. In analyses of the interaction between VAT and various SAT measures (ie, ASAT, triceps skinfold thickness, and LEG fat mass) on the odds of the MS, we evaluated these measures as continuous variables in logistic regression models. Then, for illustrative purposes, we created tertiles to plot the probability of the MS in subjects with high (top tertile) compared with lower (bottom 2 tertiles) subcutaneous adiposity across all levels of VAT mass. PROC GENMOD was used to compare odds ratios (ORs) of the MS in subjects with high and lower subcutaneous adiposity at particular levels of VAT mass.

Differences in the association between VAT area measured at different single image locations across the abdomen and at different cardiometabolic risk factor levels were tested by comparing each age-adjusted Pearson correlation coefficient with the highest correlation coefficient for each risk factor with the use of the Hotelling-Williams test (34). Then, using sex-specific logistic regression models adjusted for age, age2, BMI, current smoking status (yes or no), physical activity level, and FFM, we compared the ORs (and 95% CIs) for VAT measured at different single image locations as a predictor of the MS. ORs were based on a 1-SD increase in each of the predictor variables. Goodness-of-fit for the logistic regression models was assessed by using the c-statistic and the Hosmer-Lemeshow test, and the Akaike Information Criterion (AIC) was used to compare the final models. Multicollinearity was tested by using the variance inflation factor, and models were excluded if the variance inflation factor was >10.

VAT patterning was examined by using the anatomical position (image number) of the maximum VAT area for each person. VAT area generally follows a curvilinear pattern across the abdominal region, and it shows a peak, or maximum VAT area, for each subject; see Demerath et al (35) for further description. The anatomical location of the maximum VAT area increased in 1-cm increments from image location no. 1 (in which case peak VAT area occurred 3 cm below L4–L5) to image location no. 22 (in which case peak VAT area occurred 18 cm above L4–L5); there were no subjects with maximum VAT area at L4–L5 + 19 cm or L4–L5 + 20 cm. Maximum VAT area location was entered into sex-stratified general linear and logistic regression models as a continuous variable alongside age, age2, VAT mass, smoking, and physical activity to test the independent effects of VAT patterning on cardiometabolic risk factors and the MS.

Statistical analyses were carried out with the use of SAS software (version 9.2; SAS Institute, Cary, NC). Two-tailed (α = 0.05) tests of significance were used.


Descriptive statistics

Characteristics of the study subjects are summarized by sex in Table 1. Mean age was 43 y; ≈70% of the subjects were over-weight (37.7% of women and 44.6% of men) or obese (31.2% of women and 27.6% of men), and expected sex differences in body composition and cardiometabolic risk factor levels were observed.

Characteristics of the subjects1

Visceral adipose tissue and abdominal subcutaneous adipose tissue as predictors of the metabolic syndrome

VAT was a better predictor of the MS than was ASAT in both sexes (model AIC lower for VAT-only than for ASAT-only models; Table 2), although CIs for the 2 ORs overlapped. When VAT and ASAT were tested simultaneously (model 3), both remained significant in men, whereas, in women, ASAT was not significant. VAT remained significant in the presence of BMI (which was not significant), showing a greater effect of VAT than of either subcutaneous or total-body adiposity on the odds of the MS. However, there was a significant interaction effect between VAT and ASAT in men (P for interaction = 0.0008), and, when this interaction term was added to the model, both VAT and ASAT main effects remained significant (Table 3). This interaction is illustrated in Figure 1; at lower levels of VAT, subjects in the top tertile of ASAT had greater odds of the MS than did those with lower ASAT, but, at higher levels of VAT, subjects in the top tertile of ASAT had lower odds of the MS than did those in the bottom 2 tertiles of ASAT. The interaction term was not statistically significant in women. When we substituted triceps skinfold thickness or LEG fat mass for ASAT mass as the measure of subcutaneous adiposity, results were similar (Table 3). To illustrate these interactions numerically, we used the models to calculate the odds of the MS in men with SAT measures in the top tertile compared with those in the bottom 2 tertiles at specific VAT values. When VAT mass was set at 2 kg (relatively low for men), the adjusted OR for the MS in men with high compared with low or moderate subcutaneous adiposity was elevated (OR: 4–8). As VAT mass increased, the risk of the MS among those with high compared with low or moderate subcutaneous adiposity dropped, becoming nonsignificant at moderately high VAT mass (eg, 5 kg) and then indicating protection by elevated SAT (OR: significantly <1.0) when VAT mass was high (eg, 7 kg) (Table 3).

Interaction between visceral adipose tissue (VAT) and abdominal subcutaneous adipose tissue (ASAT) mass (by tertile) for the probability of the metabolic syndrome in men (probit estimates), n = 363.
Comparison of abdominal visceral adipose tissue (VAT), abdominal subcutaneous adipose tissue (ASAT), and BMI as predictors of the presence of the metabolic syndrome1
Interaction of visceral adipose tissue (VAT) with different subcutaneous adipose tissue (SAT) measures for the odds of the metabolic syndrome (MS) in 363 men: differential effects of elevated SAT on the odds of the MS, depending on VAT mass1

Effect of image location on associations between visceral adipose tissue and cardiometabolic risk factors

The correlation coefficients between single-image VAT areas and individual cardiometabolic risk factors are shown in Table 4. The single VAT areas having the highest correlation with each risk factor are indicated in the table by underlined r values, and the correlations that are not significantly lower than the highest correlation are indicated by boldface values. In no case was the correlation between total VAT mass and risk factor levels significantly stronger than that using the best single-image VAT area, and no single image was universally better correlated with health risks than all other images for all risk factors. Rather, a range of single images provided equivalent correlations between VAT area and cardiometabolic risk factor levels. This “optimal” range of images (identified in Table 4 by the shaded region) sometimes included the L4–L5 image and sometimes did not, depending on sex and the particular risk factor considered. L4–L5 was included in the set of optimal images for 3 of 7 risk factors in men and 5 of 7 risk factors in women. Except for LDL, images located 3–10 cm above L4–L5 in men and 2–8 cm above L4–L5 in women provided similarly high correlations between VAT area and all cardiometabolic risk factor levels. Correlations between VAT area and LDL were generally low, and they tended to reach their maximum at a higher level of the abdomen (L4–L5 + 12 cm in men and L4–L5 + 18 cm in women) than the other risk factors examined. ASAT areas across the abdomen also were compared against one another, but all were similarly correlated with the risk factors (data not shown).

Partial correlations between single visceral adipose tissue (VAT) areas and cardiometabolic risk factor levels1

Magnetic resonance imaging image location and variation in the prediction of the metabolic syndrome

There was a curvilinear relation between VAT area image location and the odds of the MS, with higher ORs in the mid-abdomen than in the upper or lower abdomen (Figure 2). The VAT image conferring the highest adjusted odds of the MS was the L4–L5 + 9 cm image in men (OR: 4.4) and the L4–L5 + 3 cm in women (OR: 3.1). Although most VAT areas were statistically equivalent to one another as predictors of the MS in both sexes (ie, the 95% CIs largely overlapped), there was a stronger pattern in men, and, in the case of the “best” predictor (L4–L5 + 9 cm) and the L4–L5 image, the 95% CIs overlapped only minimally. The relation between ASAT area and the odds of the MS did not vary by measurement location in either sex, and it was uniformly lower in magnitude (OR: ≈1.9–2.6) (data not shown).

Odds of the metabolic syndrome (National Cholesterol Education Program definition) associated with a 1-SD increase in visceral adipose tissue (VAT) area at twenty-four 1-cm-thick contiguous image locations spanning from 3 cm below L4–L5 (L4–L5 ...

Visceral adipose tissue patterns and their relation to the metabolic syndrome and related risk factors

It has been suggested that persons deposit VAT in different locations depending on their degree of adiposity and that the spatial distribution of VAT within the abdomen may have implications for disease risk. We found positive correlations between the location of the maximum VAT area and age, adiposity variables, and LDL-cholesterol concentrations. Specifically, subjects whose maximum VAT area occurred higher in the abdomen (ie more cranial than caudal) tended to be older (r = 0.20, P < 0.0001 in men only) and to have higher BMI (r = 0.24, P < 0.0001 in men; r = 0.16, P = 0.002 in women), higher VAT mass (r = 0.26, P < 0.0001 in men; r = 0.19, P = 0.0007 in women), higher ASAT mass (r = 0.33, P < 0.0001 in men; r = 0.21, P < 0.0001 in women), and higher LDL (r = 0.25, P < 0.0001 in men; r = 0.10, P = 0.04 in women). The effect of maximum VAT location on LDL-cholesterol concentrations remained significant after adjustment for age, VAT mass, FFM, smoking, and physical activity (Table 5; Figure 3); the predicted LDL-cholesterol concentration was >40 points higher in men with maximum VAT area at L4–L5 + 19 cm than in those with maximum VAT area at L4–L5 −3 cm. There was no effect of VAT patterning on the presence or absence of the MS.

Relation of visceral adipose tissue (VAT) patterning (location of maximum VAT area) to LDL-cholesterol concentrations in men (n = 363). Individual predicted values from a multivariate-adjusted regression model also included age, VAT mass, smoking status, ...
Associations between visceral adipose tissue (VAT) patterning, cardiometabolic risk factor levels, and odds of the metabolic syndrome1


The present study aimed to clarify the relation of different central and peripheral adipose tissue measures to the presence of the MS and related cardiometabolic risk factors in a large sample of essentially healthy adults. VAT mass was a strong independent predictor of the MS in men and women, even after adjustment for ASAT and BMI as well as age, smoking status, FFM, and physical activity level. This finding is consistent with previous reports, which have found that VAT is a stronger predictor of the MS and its component risk factors than is ASAT (16, 23, 26). It also is consistent with the extensive literature showing that VAT is more lipolytically active and less responsive to the adipogenic effects of insulin and that it produces a number of adipokines (eg, adiponectin, plasminogen activator inhibitor-1, and interleukin-6) in greater concentration than does SAT (7, 9, 14, 3638).

We hypothesized that greater ASAT mass, LEG fat mass, and triceps skinfold thickness may have protective effects (ie, may be associated with lower odds of theMS) in this sample of healthy adults. Goodpaster et al (16) showed in the Health, Aging, and Body Composition Study that older adults with the MS tend to have lower percentage body fat than those without the MS. In the present study, a significant interaction between VAT mass and subcutaneous adiposity traits was found in men: elevated ASAT mass and peripheral subcutaneous fat measures were deleterious when VAT mass was low, but they became more protective against the MS (ie, they reduced the OR) as VAT mass increased. There were relatively few subjects with discordant values on these measures [eg, only 4% of men were in the top tertile for VAT mass (>4.2 kg) and the bottom tertile for ASAT mass (<2.2 kg)]. The relative rarity of such subjects most likely led to the relatively wide CIs of our probability estimates at the tails of the distribution of VAT, and it may explain why smaller studies have not identified this protective effect at all. Nonetheless, our conclusions are based primarily on the significance of the interaction term in the continuous-variable models, and neither the figures nor the tables show estimates that extended past actually observed data points. Our finding is consistent with the paradigm of ectopic fat accumulation, in which increased metabolic risk stems from exceeding the capacity of the subcutaneous adipocytes to differentiate and accommodate excess circulating lipids, which results in ectopic deposition of lipids—that is, deposition in the liver, muscles, pericardium, and visceral compartments (1, 17, 18); this outcome was reviewed by Bays et al (39). The present study is, to our knowledge, the first to suggest that elevated abdominal SAT mass, as well as peripheral or extremity SAT, may have protective effects in the presence of high VAT.

Kuk et al (23) found that, whereas VAT area was a significant predictor of the MS at all 8 measurement locations they examined, the odds were twice as high for VAT area measured in the upper abdomen [near L1–L2 (unadjusted OR = 8.8)] than for VAT area measured at L4–L5 (unadjusted OR: 3.9; confidence limits not provided), which indicates that the single MRI or computed tomography image chosen for analysis could affect the perceived severity of risk associated with elevated VAT. Shen et al (27) reported that the L4–L5 image never had the highest correlation with MS risk factors and often had a significantly lower correlation than did total VAT volume or images located superior to L4–L5. They suggested choosing a single image 10–15 cm above L4–L5 in men and either 5 cm above or 5 cm below L4–L5 in women. In the present study, the wide CIs around the estimates precluded our choosing any one particular image as statistically superior to the others, but the highest risk was associated with midabdominal images (near L2–L3 in men and near L3–L4 in women), rather than with lower abdominal images (near L4–L5). By looking at individual risk factors, we identified a set of contiguous images containing the statistically equivalent, highest correlations with risk factor levels for each sex, and these ranges included the L4–L5 image ≈50% of the time. In contrast, images located 4–8 cm above L4–L5 provided correlations that were consistently high in magnitude for every risk factor examined (other than LDL), in both men and women. Therefore, we would recommend that, if a small number of images are to be analyzed, an image located in that region [which we have found to lie close to the L2–L3 intervertebral space (35)], rather than the L4–L5 image, would best capture the relation between VAT and cardiometabolic risk factors in both sexes.

Related to this finding is the documented metabolic heterogeneity within the VAT depot itself; upper abdominal VAT primarily comprises the highly lipolytic mesenteric and omental adipose tissues, which are drained by the portal vein, whereas lower abdominal VAT has a greater proportion of retroperitoneal adipose tissue, which is drained nonportally (4043). Given these differences, it is possible that not only the amount but also the spatial distribution of VAT within the abdomen has consequences for cardiometabolic risk. Novel findings from the present study are that persons vary greatly in the patterning of VAT (represented in the present analysis by the anatomical location of the maximum VAT area for each subject), that the location of the maximum VAT area tends to move upward (ie, cranially) in the abdomen with increasing age and adiposity, and that LDL-cholesterol concentrations are related to VAT patterning in men and women, independent of age and adiposity. However, the lack of an effect of VAT patterning on the odds of the MS and other risk factors indicates the need for further analysis and replication from studies with multiple-image VAT data.

A limitation of the present study is that risk factor data were available only for non-Hispanic whites; the same relations between VAT and disease risk factors may not apply in other racial-ethnic groups. Furthermore, the subjects were essentially a convenience sample including all nondiabetic subjects in 3 ongoing studies of body composition and CVD risk factors in the greater Dayton area. It is possible that the nonrandom nature of the sampling design (ie, it was restricted to relatively healthy volunteers) may have biased the relations between different adiposity measures and cardiometabolic risk. However, our results are similar to those of other studies documenting the greater risk associated with VAT than with SAT, as well as the effects of image location on cardiometabolic risk factor levels (23, 27). In addition, we adjusted for important possible confounders of the disease risk–body composition relation, including physical activity level and cigarette smoking.

In conclusion, although VAT mass was a stronger predictor of the MS than was ASAT, elevated ASAT mass (and other subcutaneous fat measures) were found to protect against the MS in men with high VAT mass, as predicted by the ectopic fat paradigm. Our findings also indicate that the L4–L5 intervertebral space is not optimal for detecting the relation of VAT to disease risk. Rather, an image located 4–8 cm cranially from L4–L5 would more strongly and consistently correlate with cardiometabolic risk factor levels. Finally, we identified a novel aspect of abdominal adiposity—the spatial patterning or topography of VAT—that appears to stem primarily from differences in age and the amount of visceral fat, but that had independent effects on LDL-cholesterol concentrations. Further work is needed to identify the genetic and environmental determinants of ectopic fat deposition and VAT patterning.


Supported by grants no. R01-DK064870, R01-DK064391, R01-HD12252, and R01-HL69995 from the National Institutes of Health.


We are grateful for the critical contributions of the Data Collection 374 staff of the Lifespan Health Research Center; the participants in the Southwest Ohio Family Heart Study, the Fels Longitudinal Study, and the Miami Valley Family Aging Study; and Carol Cottom and Kimberly Ritter for their analysis of the magnetic resonance imaging images.


1. Després J, Lemieux I. Abdominal obesity and metabolic syndrome. Nature. 2006;444:881–7. [PubMed]
2. Boyko EJ, Fujimoto WY, Leonetti DL, Newell-Morris L. Visceral adiposity and risk of type 2 diabetes: a prospective study among Japanese Americans. Diabetes Care. 2000;23:465–71. [PubMed]
3. Cefalu W, Wang ZM, Webel S, et al. Contribution of visceral fat mass to the insulin resistance of aging. Metabolism. 1995;44:954–9. [PubMed]
4. Hayashi T, Boyko EJ, Leonetti DL, et al. Visceral adiposity is an independent predictor of incident hypertension in Japanese Americans. Ann Intern Med. 2004;140:992–1000. [PubMed]
5. Kuk JL, Katzmarzyk PT, Nichaman MZ, Chruch TS, Blair SN, Ross R. Visceral fat is an independent predictor of all-cause mortality in men. Obesity. 2006;14:336–41. [PubMed]
6. Masuzaki H, Paterson J, Shinyama H, et al. A transgenic model of visceral obesity and the metabolic syndrome. Science. 2001;294:2166–70. [PubMed]
7. Misra A, Vikram NK. Clinical and pathophsyiological consequences of abdominal adiposity and abdominal adipose tissue depots. Nutrition. 2003;19:457–66. [PubMed]
8. Miyazaki Y, Glass L, Triplitt C, Wajcberg E, Mandarino L, DeFronzo RA. Abdominal fat distribution and peripheral and hepatic insulin resistnace in type 2 diabetes mellitus. Am J Physiol Endocrinol Metab. 2002;283:E1135–43. [PubMed]
9. Montague CT, O'Rahilly S. The perils of portliness: causes and consequences of visceral adiposity. Diabetes. 2000;49:883–8. [PubMed]
10. Nicklas B, Penninx B, Cesari M, Kritchevsky S, Newman A, Kanaya A. Association of visceral adipose tissue with incident myocardial infarction in older men and women: the Health, Aging and Body Composition Study. Am J Epidemiol. 2004;160:741–9. [PubMed]
11. Motoshima H, Wu X, Sinha MK, et al. Differential regulation of adiponectin secretion from cultured human omental and subcutaneous adipocytes: effects of insulin and rosiglitazone. J Clin Endocrinol Metab. 2002;87:5662–7. [PubMed]
12. Nguyen-Duy TB, Nichaman MZ, Church TS, Blair SN, Ross R. Visceral fat and liver fat are independent predictors of metabolic risk factors in men. Am J Physiol Endocrinol Metab. 2003;284:E1065–71. [PubMed]
13. Wajchenberg BL. Subcutaneous and visceral adipose tissue: their relation to the metabolic syndrome. Endocr Rev. 2000;21:697–738. [PubMed]
14. Wajchenberg BL, Giannella-Neto D, da Silva ME, Santos RF. Depot-specific hormonal characteristics of subcutaneous and visceral adipose tissue and their relation to the metabolic syndrome. Horm Metab Res. 2002;34:616–21. [PubMed]
15. Freedland E. Role of a critical visceral adipose tissue threshold (CVATT) in metabolic syndrome: implications for controlling dietary carbohydrates: a review. Nutr Metab (Lond) 2004;1:12. Internet: (accessed 15 May 2008) [PMC free article] [PubMed]
16. Goodpaster BH, Krishnaswami S, Harris TB, et al. Obesity, Regional body fat distribution, and the metabolic syndrome in older men and women. Arch Intern Med. 2005;165:777–83. [PubMed]
17. Rosouli N, Molavi B, Elbein S, Kern P. Ectopic fat accumulation and metabolic syndrome. Diabetes Obes Metab. 2007;9:1–10. [PubMed]
18. Danforth E., Jr. Failure of adipocyte differentiation causes type II diabetes mellitus? Nat Genet. 2000;26:13. [PubMed]
19. Weber R, Buckley M, Fried S, Kral J. Subcutaneous lipectomy causes a metabolic syndrome in hamsters. Am J Physiol Regul Integr Comp Physiol. 2000;279:R936–43. [PubMed]
20. Gavrilova O, Marcus-Samuels B, Graham D, et al. Surgical implantation of adipose tissue reverses diabetes in lipoatrophic mice. J Clin Invest. 2000;105:271–8. [PMC free article] [PubMed]
21. Shen W, Wang ZM, Punyanita M, et al. Adipose tissue quantification by imaging methods: a proposed classification. Obes Res. 2003;11:5–16. [PMC free article] [PubMed]
22. Demerath E, Shen W, Lee M, et al. Approximation of total visceral adipose tissue with a single magnetic resonance image. Am J Clin Nutr. 2007;85:362–8. [PMC free article] [PubMed]
23. Kuk JL, Church TS, Blair SN, Ross R. Does measurement site for visceral and abdominal subcutaneous adipose tissue alter associations with the metabolic syndrome? Diabetes Care. 2006;29:679–84. [PubMed]
24. Lee S, Janssen I, Ross R. Interindividual variation in abdominal subcutaneous and visceral adipose tissue: influence of measurement site. J Appl Physiol. 2004;97:948–54. [PubMed]
25. Shen W, Punyanitya M, Wang ZM, et al. Visceral adipose tissue: relations between single-slice areas and total volume. Am J Clin Nutr. 2004;80:271–8. [PMC free article] [PubMed]
26. Carr D, Utzschneider K, Hull R, et al. Intra-abdominal fat is a major determinant of the National Cholesterol Education Program Adult Treatment Panel III criteria for the metabolic syndrome. Diabetes. 2004;53:2087–94. [PubMed]
27. Shen W, Punyanitya M, Chen J, et al. Visceral adipose tissue: relationships between single slice areas at different locations and obesity-related health risks. Int J Obes (Lond) 2007;31:763–9. [PMC free article] [PubMed]
28. Machann J, Thamer C, Schnoedt B, et al. Standardized assessment of whole body adipose tissue topography by MRI. J Magn Reson Imaging. 2005;21:455–62. [PubMed]
29. Demerath E, Ritter K, Couch W, et al. Validity of a new automated software program for visceral adipose tissue estimation. Int J Obes (Lond) 2007;31:285–91. [PMC free article] [PubMed]
30. Levy J, Matthews D, Hermans M. Correct homeostasis model assessment (HOMA) evaluation uses the computer program. Diabetes Care. 1998;21:2191–2. [PubMed]
31. Grundy S, Cleeman J, Daniels S, et al. Diagnosis and management of the metabolic syndrome: an American Heart Association/National Heart, Lung, and Blood Institute Scientific Statement. Circulation. 2005;112:2735–52. [PubMed]
32. Lumley T, Emerson S, Chen L. The importance of the normality assumption in large public health datasets. Annu Rev Public Health. 2002;23:151–69. [PubMed]
33. Baecke JA, Burema J, Frijters JE. A short questionnaire for the measurement of habitual physical activity in epidemiological studies. Am J Clin Nutr. 1982;36:936–42. [PubMed]
34. Steiger JH. Tests for comparing elements of a correlation matrix. Psychol Bull. 1980;87:245–61.
35. Demerath E, Sun S, Rogers N, et al. The anatomical patterning of visceral adipose tissue: race, sex, and age variation. Obesity. 15:2984–93. [PMC free article] [PubMed]
36. Fain J, Madan A, Hiler M, Cheema P, Bahouth S. Comparison of the release of adipokines by adipose tissue, adipose tissue matrix, and adipocytes from visceral and subcutaneous abdominal adipose tissues of obese humans. Endocrinology. 2004;145:2273–82. [PubMed]
37. Fisher F, McTernan P, Valsamakis G, et al. Differences in adiponectin protein expression: effect of fat depots and type 2 diabetes status. Horm Metab Res. 2002;34:650–4. [PubMed]
38. Pantanetti P, Garrapa G, Mantero F, Boscaro M, Falioia E, Venarucci D. Adipose tissue as an endocrine organ? A review of recent data related to cardiovascular complications of endocrine dysfunctions. Clin Exp Hypertens. 2004;26:387–98. [PubMed]
39. Bays H, Mandarino L, DeFronzo R. Role of the adipocyte, free fatty acids, and ectopic fat in pathogenesis of type 2 diabetes mellitus: peroxismal proliferator-activated receptor agonists provide a rational therapeutic approach. J Clin Endocrinol Metab. 2004;89:463–78. [PubMed]
40. Rebuffe-Scrive M, Anderson B, Olbe L, Bjorntorp P. Metabolism of adipose tissue in intraabdominal depots in severely obese men and women. Metabolism. 1990;39:1021–5. [PubMed]
41. Bergman RN, Kim SP, Catalano KJ, et al. Why visceral fat is bad: mechanisms of the metabolic syndrome. Obesity. 2006;14:16S–9S. [PubMed]
42. Kabir M, Catalano KJ, Ananthnarayan S, et al. Molecular evidence supporting the portal theory: a causative link between visceral adiposity and hepatic insulin resistance. Am J Physiol Endocrinol Metab. 2005;288:E454–61. [PubMed]
43. Ross R, Janssen I. Computed tomography and magnetic resonance imaging. In: Heymsfield SB, Lohman TG, Wang ZM, Going SB, editors. Human body composition. 2nd ed. Human Kinetics; Champaign, IL: 2004. pp. 89–108.