Impact of Race/Ethnicity on the Relationship Between Visceral Fat and Inflammatory Biomarkers




The purpose of this study was to determine whether racial/ethnic differences exist in the relationship between visceral adipose tissue (VAT) and selected inflammatory biomarkers. Subjects included 136 African-American, 133 Hispanic, and 100 white men and women, aged ≥45. Waist circumference and BMI were measured using standard methods. Total VAT, and VAT and subcutaneous adipose tissue (SAT) at the L4L5 spinal level were measured using computed tomography. Interleukin-6 (IL-6), C-reactive protein (CRP), and fibrinogen were measured from fasting blood samples. Results revealed that waist circumference and BMI were similar among groups but African Americans had significantly lower L4L5 VAT compared with Hispanics and whites. Despite lower VAT, African-American men had similar concentrations of inflammatory biomarkers. On the other hand, African-American women had higher CRP and IL-6 than white women, and higher fibrinogen than both Hispanic and white women. After controlling for L4L5 VAT, L4L5 SAT, and age, African-American women had higher concentrations of IL-6 and fibrinogen. Stratified analyses for CRP indicated that L4L5 SAT was associated with CRP in African-American and white women after controlling for L4L5 VAT and age, but that the reverse was not true. These data indicate that African Americans had lower VAT but similar or higher concentrations of inflammatory biomarkers. African-American women consistently displayed greater inflammation compared with whites, even after controlling for VAT or SAT.


Emerging evidence suggests that there are racial/ethnic differences in relationships between BMI or waist circumference and percent body fat or visceral adipose tissue (VAT). Much research has focused on comparisons between those of European/white background and those of South Asian origin (1,2,3,4). Similar disparities in relationships between anthropometric measurements and body fat have been demonstrated in racial/ethnic groups in the United States, particularly in African Americans. Several studies have reported that African-American women had lower amounts of VAT for a given waist circumference, BMI, or waist-to-hip ratio compared with white women (5,6,7,8,9). Less research has focused on Hispanics. However, one small study found that external circumference measures were poor predictors of VAT in Mexican-American women (10). Unfortunately, this study suffered from lack of comparison to any other group.

We observed that middle-aged and older African-American men and women had similar BMI and waist circumference measurements compared with whites and Hispanics, but had lower total VAT and VAT measured at the L4L5 spinal level (L4L5 VAT) (11). If VAT is the underlying culprit contributing to metabolic risk, lower VAT in African Americans would be hypothesized to result in lower cardiovascular risk related to adipose tissue by-products. However, African Americans (12) and/or Hispanics (13) have been shown to have higher concentrations of inflammatory markers, such as C-reactive protein (CRP) and fibrinogen. Such chronic low-grade inflammation may contribute to increased cardiovascular disease risk in these populations. These disparate observations whereby African Americans and/or Hispanics exhibit the same or lower VAT but higher inflammation have not been reconciled. However, subcutaneous adipose tissue (SAT) has been reported to be more strongly associated with metabolic disturbances such as insulin resistance in African Americans (9), suggesting that overall body fatness remains important in mediating disease processes. Therefore, the purpose of this study was to determine whether independent racial/ethnic differences in the relationship between VAT and selected inflammatory biomarkers remained after accounting for SAT among middle-aged and older adults.

Methods and Procedures


Subjects were recruited from North Texas Primary Care Practice-Based Research Network physician practices, and from advertisements in the Fort Worth, TX, area. Subjects were ≥45 years of age and had no current or previous history of cardiovascular disease. Race/ethnicity was obtained by self-report. No selection was made on body type or degree of overweight. The cohort of 369 included 136 African-American, 133 Hispanic, and 100 white men and women. Group anthropometric characteristics are presented using this cohort. Of these, usable computed tomography (CT) images of VAT were obtained from 121 African Americans, 126 Hispanics, and 95 whites. Group averages of VAT used this smaller cohort. Because some subjects' body habitus exceeded the field of view of the CT scanner, the entire SAT area was not always visible on the L4L5 scan. As a result, a subset of 109 African Americans, 117 Hispanics, and 86 whites was used in L4L5 SAT analyses. All procedures were approved by the Institutional Review Boards at the University of North Texas Health Science Center and John Peter Smith Hospital, and by the Community Advisory and Scientific Review Boards of the North Texas Primary Care Practice-Based Research Network. Informed consent was obtained from all subjects.

Data from medical history questionnaires revealed that 15.6 and 9.2% of Hispanic men and women, 22.2 and 9.5% of African-American men and women, and 13.2 and 16.7% of white men and women, respectively, were current smokers. Questionnaire data also revealed that lipid-lowering medications were taken by 30.3 and 21.5% of Hispanic men and women, 27.0 and 31.3% of African-American men and women, and 28.9 and 33.3% of white men and women, respectively. In addition, nonsteroidal anti-inflammatory drugs (NSAIDs) were taken by 3.0 and 13.9% of Hispanic men and women, 21.6 and 21.3% of African-American men and women, and 13.2 and 20.8% of white men and women, respectively.

Anthropometric measures of body composition

Height was measured to the nearest 0.25 inch and weight was measured to the nearest 0.25 lb using a standard balance scale. Waist and hip circumferences were measured to the nearest 1/16 inch using a nonstretchable tape. Measurements were converted to appropriate SI units. Waist circumference was measured midway between the bottom of the ribcage and the top of the iliac crest. Hip circumference was measured at the level of maximum size of the buttocks (14). Percent body fat was determined using bioimpedance (Tanita Model TBF-300; Tanita, Arlington Heights, IL). Waist-to-hip and waist-to-height ratios were calculated.

CT-determined VAT and SAT

Abdominal adipose tissue was measured using a 16-slice CT scanner (Toshiba Aquilion 16, Model #TSX-101A; Toshiba America Medical Systems, Tustin, CA). Eight axial CT images of the abdomen were obtained from each subject, with slice number six centered on the L4L5 interspace. An additional five slices were obtained every 5 cm above the L4L5 interspace, and two additional slices were obtained 5 and 10 cm below the L4L5 interspace. Collimation of 8 mm was used for each axial slice. The cross-sectional areas of VAT were quantified on each slice using software-derived algorithms (Analyze, version 6.0; Biomedical Imaging Resource, Rochester, MN). SAT was calculated from each slice as previously described (11). From the combined measurements of the eight slices, total abdominal VAT and SAT masses were calculated according to previously published methods (15).

Laboratory analyses

Fasting venous blood samples were taken to measure CRP, interleukin-6 (IL-6), and fibrinogen concentrations. Samples from subjects with active infections were not used in analyses of these measures. Likewise, CRP concentrations >10 were not used in any analyses. Samples were analyzed by a commercial laboratory (Quest Diagnostics, Dallas, TX).

Statistical analyses

All analyses were performed separately for men and women. Racial/ethnic anthropometric characteristics are presented as mean ± s.d. One-way ANOVA with Tukey–Kramer's post hoc test was used to determine differences among racial/ethnic groups. The Tukey–Kramer test was used to account for unequal group sizes. Inflammatory biomarker concentrations were examined for normality. IL-6 was log transformed before analysis and compared using one-way ANOVA with Tukey–Kramer's post hoc test. Due to violations of normality, CRP and fibrinogen are presented as medians and interquartile ranges, and were analyzed using the Kruskal–Wallis test. Multiple linear regression analysis was used to determine if relationships between L4L5 VAT and inflammatory biomarkers (IL-6, CRP, and fibrinogen) were modified by race/ethnicity (L4L5 VAT × race/ethnicity interaction) after controlling for SAT and age. Statistical significance was accepted at the 0.05 level.


Anthropometric characteristics and VAT

Racial/ethnic group characteristics for men and women are shown in Table 1. Among both men and women, the three groups were similar in age. However, both white and African-American men and women were taller than their Hispanic counterparts (both P ≤ 0.05). Among women, African Americans were significantly heavier than Hispanics, and had a higher body fat percentage then whites (both P ≤ 0.05). The waist-to-height ratio was greater in Hispanic women compared with white counterparts (P ≤ 0.05). Importantly, BMI, waist circumference, waist-to-hip ratio, and sagittal diameter did not differ among groups in either men or women.

Table 1.  Anthropometric characteristics by race/ethnicity and gender
inline image

Despite similar average waist circumference and BMI measurements, L4L5 VAT was significantly lower in African-American men and women compared with both Hispanic and white men and women (Figure 1a). Similarly, among men, total VAT was lower in African Americans compared with both Hispanics and whites. However, among women, total VAT was lower in both whites and African Americans compared with Hispanics (Figure 1b), but there was no difference between whites and African Americans.

Figure 1.

Computed tomography results. (a) L4L5 VAT, (b) total VAT, and (c) L4L5 VAT/SAT ratio in white, African-American and Hispanic men and women. Data are mean ± s.d. *P ≤ 0.0001, greater than African Americans of the same gender; **P ≤ 0.05, greater than whites of the same gender. SAT, subcutaneous adipose tissue; VAT, visceral adipose tissue.


L4L5 SAT did not differ among men or among women (Table 1). As a result, there were racial/ethnic differences in the proportion of VAT and SAT at L4L5. The L4L5 VAT/SAT ratio was significantly higher in Hispanic and white men compared with African-American men (P ≤ 0.01; Figure 1c). Among women, the L4L5 VAT/SAT ratio was also higher in Hispanics compared with whites and African Americans, and was also higher in whites compared with African Americans (P ≤ 0.0001). In both men and women, the lower ratio among African Americans was due to similar L4L5 SAT but lower L4L5 VAT measurements.

Inflammatory biomarker concentrations

Despite lower VAT in African Americans, African-American and Hispanic women had higher IL-6 concentrations compared with white women (Table 2; P ≤ 0.001). African-American women also had higher CRP concentrations (P ≤ 0.05) compared with white women, and higher fibrinogen concentrations compared with both white and Hispanic women (P ≤ 0.01). Among men, there tended to be group differences in IL-6 concentrations (P = 0.09), with Hispanics tending to have greater concentrations than whites. There were no significant racial/ethnic differences among men in CRP or fibrinogen (P > 0.05).

Table 2.  Plasma concentrations of inflammatory risk markers by race/ethnicity and gender
inline image

Relationships among VAT, SAT, and inflammatory biomarkers

Linear relationship between L4L5 VAT and inflammatory markers is shown in Figure 2. Results from multiple linear regression analyses are shown in Table 3. Significant associations between race and IL-6, between L4L5 VAT and IL-6, and between L4L5 SAT and IL-6 were observed. Among women, African Americans (P = 0.06) and Hispanics (P = 0.09) tended to have a higher IL-6 than whites. Among men, Hispanics had a higher IL-6 than whites (P = 0.03), but differences were not observed between African-American and white men. Additionally, while controlling for race/ethnicity, L4L5 SAT, and age, increases in L4L5 VAT were independently associated with increases in IL-6 in both men (P = 0.03) and women (P = 0.0032). Further, after controlling for race/ethnicity, VAT, and age, increases in L4L5 SAT were also independently associated with increases in IL-6 in men (P = 0.02) and women (P = 0.0001).

Figure 2.

VAT and inflammation. Linear relationship between L4L5 VAT and (a) ln IL-6 in women, (b) ln IL-6 in men, (c) CRP in women, (d) CRP in men, (e) fibrinogen in women, and (f) fibrinogen in men. CRP, C-reactive protein; IL-6, interleukin-6; VAT, visceral adipose tissue.

Table 3.  Results of multiple linear regression models with race/ethnicity, L4L5 VAT, L4L5 SAT and age as independent variables
inline image

In analyses for CRP, no association between race and CRP was observed in men, but significant associations between L4L5 VAT and CRP, and between L4L5 SAT and CRP were observed (Table 3). For every cm2 increase in L4L5 VAT, CRP increased by 0.007 mg/l, while for every cm2 increase in L4L5 SAT, CRP increased by 0.006 mg/l. In women, there was a significant interaction between race and L4L5 SAT when predicting CRP (P = 0.05). Therefore, regression analyses were stratified by race/ethnicity for women (Table 4). While controlling for SAT and age, there was no significant association between VAT and CRP in any group. However, while controlling for VAT and age, there were significant associations between SAT and CRP for whites (P = 0.05) and African Americans (P = 0.0005). For every cm2 increase in L4L5 SAT, there was a 0.006 mg/dl increase in CRP in white women and a 0.011 mg/dl increase in CRP in African-American women.

Table 4.  Relationships among VAT, SAT, age, and CRP in women
inline image

No associations between being Hispanics or African Americans (compared to white) and fibrinogen were observed among men (Table 3). Additionally, L4L5 VAT was not associated with fibrinogen among men, but there was an association between L4L5 SAT and fibrinogen (P = 0.003). Every cm2 increase in L4L5 SAT was associated with a 0.19 mg/dl increase in fibrinogen in men. Among women, however, African Americans had higher fibrinogen concentrations than whites (P = 0.01). Also, while controlling for race/ethnicity, L4L5 VAT, and age, L4L5 SAT had a significant association with fibrinogen in women: every cm2 increase in L4L5 SAT was associated with a 0.18 mg/dl increase in fibrinogen (P = 0.0002). Finally, while controlling for race/ethnicity, L4L5 VAT, and L4L5 SAT, age had a significant association with fibrinogen in women. Every year increase in age was associated with a 1.76 mg/dl increase in fibrinogen (P = 0.01).


There were several significant findings in this study. First, African-American men and women had either similar or greater concentrations of selected inflammatory biomarkers compared with whites and Hispanics, despite having lower VAT. Significant associations between race/ethnicity and inflammatory biomarkers were observed more in women than in men. In particular, African-American women consistently exhibited greater concentrations of the selected inflammatory biomarkers compared with white women after controlling for L4L5 VAT, L4L5 SAT, and age. In contrast, the only association among men between race/ethnicity and inflammatory biomarkers was a higher IL-6 among Hispanics compared with whites. Finally, despite the presumed relationship of VAT to various disease processes, L4L5 SAT was more consistently associated with inflammatory biomarker concentrations after controlling for L4L5 VAT and age in both men and women, but the reverse was not always true. This was particularly true for fibrinogen, and for CRP in women. Taken together these findings suggest that there are racial/ethnic differences among middle-aged and older adults regarding the influence of body fat amounts and/or distribution on selected plasma inflammatory biomarker concentrations.

The present findings lend weight to other studies showing that young African-American women have lower VAT than whites with the same BMI, waist circumference or waist-to-hip ratio (5,6,7,8,9), and extend the findings to middle-aged and older women and men. However, lower VAT in African Americans would be hypothesized to be associated with lower concentrations of inflammatory biomarkers. Contrary to this hypothesis, African Americans had either similar or greater plasma concentrations of selected inflammatory biomarkers. These data may have important implications for increased rates of inflammation-related diseases, such insulin resistance or diabetes (16), particularly among women.

Few studies have measured both VAT and inflammatory biomarker concentrations in a multiethnic population. One small study comparing European and South Asian men and women found that both CRP and VAT were higher in South Asian women compared with their European counterparts even at the same BMI (17). Further, CRP was more strongly related to VAT in South Asians, suggestive of a different VAT–CRP relationship in this population.

Similarly, there are only sparse data on inflammatory biomarker concentrations in multiethnic groups in the US. These studies suggested that fibrinogen may be higher in African Americans (12,18,19), although not all studies showed this trend (20,21). CRP concentrations were also higher in African Americans than in whites in several large studies (12,22,23), and were correlated with anthropometric measures of obesity (22). In particular, data from the Women's Health Study (23) showed higher CRP in African-American women compared with other groups. In contrast, NHANES data (21) did not show this trend and found instead that Mexican-American women had higher CRP compared with whites. However, none of these studies measured VAT, so that potential ethnic differences in the VAT–inflammatory biomarker relationship could not be ascertained. In one study that measured VAT, the negative association between VAT and adiponectin was stronger in African Americans (24), suggesting that African Americans may suffer greater consequences from VAT accumulation.

This study provided some support for this idea, showing that African Americans had similar or greater inflammatory biomarker concentrations at lower amounts of VAT. However, overall body fatness may still play a role in inflammation because SAT also had significant independent associations with inflammatory biomarker concentrations. This may explain in part the independent association of CRP and waist circumference shown among women in NHANES (21). However, African-American women in this study consistently exhibited greater concentrations of the selected inflammatory biomarkers even after controlling for both L4L5 VAT and SAT. Further, the higher concentrations of inflammatory biomarkers among African-American women occurred despite similar or lower self-reported rates of smoking, and similar or higher self-reported rates of taking lipid-lowering medications or nonsteroidal anti-inflammatory drugs. Together, this suggests a higher basal level of inflammation but similar inflammatory responses to increasing amounts of body fat among African-American women. The cause of the higher basal inflammation can only be speculated upon but possibilities include higher intrinsic activity of cytokine pathways and/or different lifestyle influences on inflammation. Thus, simply measuring inflammatory biomarker concentrations may obscure important racial/ethnic differences in the influence of VAT on inflammation, and by extension, disease.

IL-6 is known to be secreted from a number of different cells, including activated macrophages and lymphocytes. A substantial portion of circulating IL-6 originates from adipose tissue, with VAT contributing more to circulating IL-6 than SAT. In turn, IL-6 has a primary role in the acute phase production of CRP by the liver (25). Fibrinogen is also produced by the liver in response to cytokine stimulation. Evidence is accumulating that macrophages infiltrate into white adipose tissue, and may contribute to cytokine production by adipose tissue (26). Yet it is not clear whether any of these factors are subject to racial/ethnic variation. One intriguing study found that upper body obesity was associated with insulin resistance at the level of the subcutaneous adipocyte in white but not African-American women (27). This in vitro study mirrored the results of an earlier in vivo study, in that upper body obesity in African-American women did not exacerbate insulin resistance to the same extent as in white women (28). These findings suggest the possibility that genetic factors affect important metabolic pathways but the study did not offer potential mechanisms.

Aside from intrinsic cytokine production pathways, lifestyle factors such as diet or exercise may play a role in the altered VAT/body fat–inflammatory biomarker relationship seen in this study. An observational study found that diets high in glycemic load were associated with increased concentrations of CRP, and that the dose–response gradient between glycemic load and CRP was more exaggerated in overweight women (29). It was speculated that recurrent postprandial hyperglycemia may lead to the production of advanced glycation end products, which may stimulate the liver to increase production of acute phase reactants such as CRP. A later prospective study subjected volunteers to diets high in either sucrose or artificial sweeteners, and found an association between glycemic load and both haptoglobin and transferrin, but not CRP (30). Other dietary factors have also been shown to influence inflammation. In particular, fats and processed meats have been associated with increased inflammation, while fiber, fruits, and vegetables have been associated with reduced inflammation (31,32). Moreover, such dietary patterns may be significantly associated with race–ethnicity. African Americans may have eating patterns reflecting higher consumption of fat and calories, and lower consumption of fruits and vegetables (33). Further, African Americans, women in particular have been shown to have lower daily energy expenditure or physical activity compared with whites (34,35,36,37), which may independently contribute to inflammation. However, whether exercise or dietary differences in this study accounted for the altered VAT–inflammatory biomarker relationships among African Americans is unknown.

The present data indicated that SAT may consistently be associated with inflammatory biomarker concentrations, even after controlling for VAT, but that the reverse was not always true. This is in partial support of other studies demonstrating that SAT was significantly related to metabolic disturbances such as insulin resistance (38,39). One study addressed possible racial/ethnic differences in this relationship and reported that SAT was significantly correlated with insulin sensitivity and fasting insulin in African-American, but not white, women (9). This suggests that total truncal fatness, and not just visceral fat, may be an important contributor to the inflammatory process in some racial/ethnic groups.

Previously, we identified that anthropometric measurements of central body adiposity, such as waist circumference and BMI, represented different amounts of VAT in African Americans (11). This study provided an additional complicating factor, demonstrating that given amounts of VAT represent different concentrations of selected inflammatory biomarkers in middle aged and older African Americans, Hispanics, and whites. Taken together, the link between anthropometric measurements, VAT, and inflammation appears to display racial/ethnic differences in this age group. Further refinements of our understanding of the role of overall body fatness and fat distribution on various risk factors are clearly warranted. In addition, anthropometric or VAT cutoff points used to define elevated risk may also need refinement in order to accommodate these altered relationships.

These data provide intriguing insights into important racial/ethnic differences between visceral fat and plasma concentrations of inflammatory biomarkers. However, results must be viewed with caution because of the relatively small number of subjects in the race/gender subgroupings. Also, the cross-sectional nature of the study does not allow for elucidation of cause-and-effect relationships. Further research is needed to determine the generalizability of these results to groups that differ in age, health status, and/or geographic location. Notably, hormone replacement status among women needs to be accounted for because this therapy has been associated with increased CRP (21,23). Determining the effects of diet or exercise on inflammation may also be warranted to further refine these findings. Finally, whether altered VAT–biomarker relationships are linked through increased disease incidence to increased morbidity or mortality remains to be determined.

This study found that African Americans had similar or higher concentrations of CRP, IL-6, and fibrinogen, despite having lower amounts of VAT. African-American women in particular displayed greater inflammatory biomarker concentrations after controlling for VAT and SAT. However, after controlling for SAT, the influence of VAT on CRP and fibrinogen was not longer significant in this group. Whether these differences are due to changes in intrinsic cytokine production pathways or to different lifestyle factors needs further study.


This study was supported by a grant P20 MD001633 from the National Center for Minority Health and Health Disparities (MCMHD), National Institutes of Health. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the NCMHD.


The authors declared no conflict of interest.