A growing body of evidence has consistently shown a correlation between obesity and chronic subclinical inflammation. It is unclear whether the size of specific adipose depots is more closely associated with concentrations of inflammatory markers than overall adiposity. This study investigated the relationship between inflammatory markers and computerized tomography-derived abdominal visceral and subcutaneous fat and thigh intermuscular and subcutaneous fat in older white and black adults. Data were from 2,651 black and white men and women aged 70–79 years participating in the Health, Aging, and Body Composition (Health ABC) study. Inflammatory markers, interleukin-6 (IL-6), C-reactive protein (CRP), and tumor necrosis factor-α (TNF-α) were obtained from serum samples. Abdominal visceral and subcutaneous fat and thigh intermuscular and subcutaneous fat were quantified on computerized tomography images. Linear regression analysis was used to evaluate the cross-sectional relationship between specific adipose depots and inflammatory markers in four race/gender groups. As expected, blacks have less visceral fat than whites and women less visceral fat than men. However, abdominal visceral adiposity was most consistently associated with significantly higher IL-6 and CRP concentrations in all race/gender groups (P < 0.05), even after controlling for general adiposity. Thigh intermuscular fat had an inconsistent but significant association with inflammation, and there was a trend toward lower inflammatory marker concentration with increasing thigh subcutaneous fat in white and black women. Despite the previously established differences in abdominal fat distribution across gender and race, visceral fat remained a significant predictor of inflammatory marker concentration across all four subgroups examined.
A consistent correlation has been shown between obesity and chronic subclinical inflammation (1,2). In particular, obesity has been associated with elevated concentrations of the inflammatory marker C-reactive protein (CRP), as well as high concentrations of the proinflammatory cytokines interleukin-6 (IL-6) and tumor necrosis factor-α (TNF-α) (1). Adipose tissue itself produces cytokines (3,4), and it has been estimated that adipose tissue is responsible for about 25% of systemic IL-6 (4). Increased concentrations of IL-6, CRP, and TNF-α have been associated with higher risk of heart failure (5), lower insulin sensitivity (6), and metabolic syndrome (7,8), among other conditions. In animal studies, the removal of visceral adipose tissue reduced the negative metabolic effects of diabetes and lowered the expression of TNF-α in subcutaneous adipose tissue (9,10).
Studies have suggested that specific body fat depots, rather than overall adiposity, may be more closely associated with concentrations of IL-6, CRP, and TNF-α (1,4,11,12). Visceral adiposity in particular has been shown to be uniquely correlated with concentrations of inflammatory markers (11,12,13). Adipose depots in the thigh may also play an important metabolic role; lower thigh subcutaneous adipose and higher thigh intermuscular adipose are associated with increased risk of metabolic problems (14,15). However, studies of the effect of body fat distribution on inflammation have been limited by a lack of detailed measurements in multiple fat depots, especially in depots outside of the abdomen.
Few studies have addressed adipose distribution and inflammatory marker concentration across racial groups (16) in older persons. There is evidence that body fat distribution differs between races; it has been shown that Asians have more abdominal visceral fat than whites, and that whites have more abdominal visceral fat than blacks (16,17,18,19). In addition, blacks have more intermuscular adipose tissue than whites; on average the amount of intermuscular adipose tissue in blacks may be almost equal to the amount of visceral adipose tissue. Still, it is not known whether the relationship between fat distribution and inflammation is similar across all races and ethnicities.
Recent findings by Pou et al. show an association between visceral abdominal fat and concentrations of CRP and IL-6 (12). However, this study cohort is almost entirely Caucasian, is younger than the Health, Aging, and Body Composition (Health ABC) cohort, and the analysis was limited to only two fat depots, both of which were in the abdomen. In contrast, the Health ABC study includes a substantial number of both white and black participants, and included computerized tomography scans of the abdomen and thigh. While the importance of the inflammation markers in this study has been demonstrated in studies of cardiovascular disease, diabetes, pneumonia, and cancer, the explicit association between adipose depots and inflammation by race and gender has not been addressed.
We hypothesized that visceral adipose tissue would be associated with inflammatory markers in white men and women but that the relationship between visceral adipose tissue and inflammation would be weaker in blacks due to smaller visceral adipose tissue depots. We also hypothesized that intermuscular adipose tissue, present in larger amounts in blacks, would also be associated with inflammation in this subgroup.
Data were from the Health ABC study, a longitudinal cohort study of body composition and health consisting of 3,075 well-functioning, 70- to 79-year-old black and white men and women. Participants were identified from a random sample of white Medicare beneficiaries and all age-eligible community-dwelling black residents in designated zip code areas surrounding Memphis, Tennessee, and Pittsburgh, Pennsylvania. Participants were eligible if they reported no difficulty in either walking one quarter of a mile, going up 10 steps without resting, or performing basic activities of daily living. Participants were excluded if they reported a history of active treatment for cancer in the prior 3 years, planned to move out of the study area in the next 3 years, or were currently participating in a randomized trial of a lifestyle intervention. Baseline data, collected between April 1997 and June 1998, included an in-person interview and a clinic-based examination, with evaluation of body composition, clinical, and subclinical diseases, and physical functioning. All participants provided written informed consent, and all protocols were approved by the institutional review boards of the University of Tennessee at Memphis, the University of Pittsburgh, and the University of California, San Francisco. Of the original Health ABC cohort of 3,075 individuals, 424 individuals were excluded due to incomplete data on inflammatory markers, body composition, or covariates, leaving 2,651 subjects for the present cross-sectional analyses. The 424 excluded persons had mean age of 74.2 ± 2.9 years and included 108 white men, 81 black men, 98 white women, and 137 black women.
Body composition. Abdomen and thigh computerized tomography scans were obtained at the Pittsburgh site using a 9800 Advantage Scanner (General Electric, Milwaukee, WI) and either a Somatom Plus 4 (Siemens, Erlangen, Germany) or a Picker PQ 2000S (Marconi Medical Systems, Cleveland, OH) scanner at the Memphis site (14). The scans were conducted at 120 kVp, 200–250 mA seconds, at a slice thickness of 10 mm. Areas were calculated by multiplying the number of pixels of a given tissue type by the pixel area using ILD development software (RSI Systems, Boulder, CO). Scans of the abdomen were taken at the level of the space between the fourth and fifth lumbar vertebrae (L4–L5). The scan at mid-thigh level was performed at one half of the distance between the medial edge of the greater trochanter and the intercondyloid fossa. Visceral fat was manually distinguished from subcutaneous fat by tracing along the fascial plane defining the internal abdominal wall. In the thighs, intermuscular and visible intramuscular fat tissue were separated from subcutaneous adipose tissue by drawing a line along the deep fascial plane surrounding the thigh muscles.
Inflammatory markers. Measures for the cytokines IL-6 and TNF-α and for CRP were obtained from frozen stored plasma or serum. Fasting blood samples were obtained in the morning, and after processing, the specimens were aliquoted into cryovials, frozen at −70 °C, and shipped to the Health ABC Core Laboratory at the University of Vermont. Cytokines were measured in duplicate by enzyme-linked immunosorbent assay kits from R&D Systems (Minneapolis, MN). The detectable limit was 0.10 pg/ml for IL-6 (by HS600 Quantikine Kit) and 0.18 pg/ml for TNF-α (by HSTA50 kit). Serum concentrations of CRP were also measured in duplicate by enzyme-linked immunosorbent assay based on purified protein and polyclonal anti-CRP antibodies (Calbiochem, San Diego, CA). The CRP assay was standardized according to the World Health Organization First International Reference Standard with a sensitivity of 0.08 µg/ml. Assays of blind duplicates collected for 150 participants showed an average interassay coefficient of variation of 10.3% for IL-6, 8.0% for CRP, and 15.8% for TNF-α. In the entire study population, the correlation coefficients between IL-6 and CRP, CRP and TNF-α, and TNF-α and IL-6 were 0.39, 0.11, and 0.21, respectively.
Covariates. Covariates were selected on the basis of previously identified associations with either obesity or inflammation. Sociodemographics included clinical site (Memphis, Pittsburgh, PA), age, marital status (never married, previously married, married), and level of education (<12 years, 12 years, >12 years). BMI was calculated as weight in kilograms divided by height in meters squared (kg/m2). Smoking was included as cigarette use in pack-years. Physical activity was measured by questionnaire in which participation in the following activities was determined: gardening or yard work, housework, stair climbing, walking for exercise, other walking, aerobics/calisthenics, weight training, high-intensity exercises, moderate-intensity exercises, or work/volunteer/caregiving activities. The intensity of walking and exercise, and the frequency and duration of each activity in the previous 7 days was ascertained. Weekly caloric expenditure was estimated in kcal/kg/h through the assignment metabolic equivalent unit values to each activity (20). Presence of heart disease, cerebrovascular disease, lung disease, peripheral arterial disease, diabetes mellitus, osteoarthritis, and cancer was determined using standardized algorithms considering self-report and use of specific medications. Depressed mood was assessed with the Center for Epidemiologic Studies Depression scale. A cutoff score of 16 was used as a criterion for major depressive symptoms (21). Cognitive impairment was defined as a Modified Mini-Mental State Examination (3MS) score <78 (22). Medications taken in the previous 2 weeks were recorded and coded according to the Iowa Drug Information System (23).
Analyses were performed separately for each of four race-gender groups because of known differences in the distribution of body adipose by race and gender. Linear regression analysis was used to evaluate the relationship between each body composition measure and serum concentrations of inflammatory markers, with inflammatory marker concentrations normalized by logarithmic transformation. Regression was performed using three different models; the first controlled for clinic site and age, the second additionally controlled for BMI to account for other body adipose that might be correlated with inflammation, and the third model controlled for all variables of the previous models and marital status, education level, pack-years of smoking, physical activity, diseases, anti-inflammatory drugs, and oral steroid medication. Reported statistical analysis values include the standard deviation increase in normalized inflammatory marker concentration per standard deviation increase in fat area, and the significance. All statistical analyses were performed using SPSS, version 14.0 (SPSS, Chicago, IL).
The main characteristics of the study population are presented by gender and race in Table 1. Concentrations of IL-6 (P < 0.01) and CRP (P < 0.01) were higher in blacks than in whites and TNF-α concentration (P < 0.01) was higher in whites than in blacks. Black women had a higher mean BMI than white women. Pack-years of smoking were higher in men than in women (P < 0.01) and in white persons than in black persons (P < 0.01). Whites had higher levels of physical activity than blacks (P < 0.01). Heart disease, lung disease, and peripheral artery disease were more prevalent in men than in women (P < 0.01). The prevalence of osteoarthritis was higher in women (P < 0.01) and a greater percentage of women had depressive symptoms than men (P = 0.01). In both men and women, blacks had a higher prevalence of diabetes mellitus (P < 0.01) and cognitive impairment (P < 0.01) while the prevalence of osteoarthritis (P < 0.01) and cancer (P < 0.01) was higher among whites. Figure 1 shows the mean abdominal and thigh fat areas by gender and race. Women had more abdominal subcutaneous fat (P < 0.01) and thigh subcutaneous fat (P < 0.01) than men. White persons had more abdominal visceral fat than black persons (P < 0.01). In both men and women, blacks had more thigh intermuscular fat (P < 0.01).
Table 1. Main characteristics of the study population
In both black and white men, visceral fat was associated with significantly higher concentrations of IL-6 and CRP but the β-coefficients did not show the association to be consistently stronger in whites or blacks, even adjusting for adiposity and other effects (Table 2). In white men, there was a significant negative association between thigh subcutaneous fat and concentrations of IL-6 and TNF-α in the fully adjusted model. Thigh intermuscular fat was related to significantly higher concentrations of IL-6 and CRP in both black and white men in model 1. After adjustment for BMI, this association remained significant for IL-6 in all men and also for CRP in black men.
Table 2. Standardized regression coefficients of inflammatory markers by body composition in white and black men
All of the body fat measures were positively and significantly associated with all three inflammatory markers in white women, and with IL-6 and CRP in black women (Table 3, model 1). Adjustment for total adiposity in model 2 attenuated most of these relationships, but abdominal visceral fat remained significantly related to higher concentrations of IL-6, CRP, and TNF-α in all three regression model in both black and white women; the strength of these associations were similar for IL-6 and TNF-α. Thigh intermuscular fat area remained significantly associated with higher concentrations of all three inflammatory markers in white women and not in black women. When the analyses in all four race/gender groups were adjusted for total percent body fat instead of BMI, the general results in both men and women do not change considerably (data not shown).
Table 3. Standardized regression coefficients of inflammatory markers by body composition in white and black women
We examined whether visceral or abdominal subcutaneous fat was more closely associated with inflammation by including both measures in a fully adjusted model (Figure 2). Visceral fat was positively associated with concentrations of all three inflammatory markers in every race/gender group. The association was significant in every case except in black men with TNF-α. In contrast, there was no consistent association between abdominal subcutaneous fat and inflammatory marker concentration after adjustment for abdominal visceral fat.
Thigh intermuscular adipose and thigh subcutaneous adipose were similarly added together to a fully adjusted model for each of the inflammation markers (Figure 3). Thigh intermuscular fat was significantly associated with higher concentrations of IL-6 in both groups of men, higher concentrations of CRP in black men, and higher concentrations of IL-6, CRP, and TNF-α in white women. There was a trend toward lower concentrations of IL-6, CRP, and TNF-α with increasing thigh subcutaneous fat, and this trend was significant for IL-6 and TNF-α in white men.
To test whether the associations with visceral adipose were secondary to diseases associated with visceral adipose rather than visceral adipose itself, we divided the participants into an “unhealthy” group and a “healthy” group (with and without diseases). The associations between inflammatory marker concentration and visceral fat remained significant in both the healthy and unhealthy group (data not shown).
We present the first examination of the association of abdominal and thigh measures of adipose depots in relation to inflammation in older black and white men and women. The main finding of this study is that in both men and women, abdominal visceral adipose area was significantly associated with higher concentrations of IL-6 and CRP across both races and these associations were independent of total adiposity. Despite systematic differences by race and gender in the amount of visceral adipose tissue, the strength of the association was similar in all four race/gender groups for IL-6 and CRP, while the association with TNF-α appeared stronger in the women, regardless of race and total adiposity. Concentrations of inflammatory markers tended to increase with greater thigh intermuscular fat, with the most consistent associations for IL-6.
The association between visceral fat and inflammation found in this study is consistent with the existing literature; however, contrary to our hypothesis this association was of similar strength in blacks and whites. Previous studies have shown that CRP concentration is significantly positively associated with visceral fat in South Asian, white, and Japanese men and women (17,24). Significant correlations between visceral fat and concentrations of IL-6 and CRP, but not TNF-α, have been documented in postmenopausal white women (25). Similarly, a large study of elderly men and women found that IL-6 and CRP, but not TNF-α, were significantly associated with visceral fat (12). In these studies, the relationships between inflammatory marker concentrations and visceral fat were independent of total adiposity measured as total percent body fat or BMI. Our findings further support the growing body of evidence that the visceral fat depot, rather than abdominal subcutaneous fat, is uniquely important to inflammation (26).
To our knowledge, no previous study has investigated the associations between inflammatory markers and thigh composition. Greater thigh intermuscular fat has been associated with type 2 diabetes and impaired glucose tolerance (15). The relationship between thigh intermuscular fat and higher inflammation, particularly for IL-6, supports evidence that thigh intermuscular adipose tissue may contribute to negative health outcomes. In addition, it has been shown that low thigh subcutaneous fat is a risk factor for a disadvantageous glucose and lipid profile (14), and that greater thigh subcutaneous fat is correlated to higher insulin sensitivity (15). Most of the coefficients suggested a trend toward lower inflammatory marker concentration with higher thigh subcutaneous fat; however, these results were inconsistent.
Several studies have demonstrated gender and race differences in visceral fat deposition. As found in this study, black men and women have been shown to have less visceral fat than white men and women (18,19). Men have been shown to have more visceral fat and less subcutaneous fat than women (27), and this comparison holds true in our study within race. However, despite these considerable differences in abdominal fat distribution across gender and race, visceral fat remained a significant predictor of inflammatory marker concentration in all analysis groups. Further studies should investigate whether visceral fat is similarly related to inflammatory markers in Asians and other racial and ethnic groups. For instance, Asian women have been shown to have significantly larger areas of visceral fat than white women (16,17).
The causality of the associations between negative health outcomes, visceral fat, and intermuscular thigh fat are not yet clear. It has been shown that the relationship between diabetes and visceral fat is attenuated by concentrations of adiponectin, IL-6, TNF-α, and plasminogen activator inhibitor 1 (28), indicating the possibility that adipocytokines are key in the link between visceral fat and an unfavorable metabolic profile. Adipose tissue itself secretes IL-6 and expresses TNF-α (4). However, it is noteworthy that this and other studies have found consistent visceral fat associations with concentrations of IL-6 and CRP, but not TNF-α (12,25). An investigation of IL-6 and TNF-α concentration in 39 healthy subjects indicated that subcutaneous adipose tissue releases IL-6, but does not secrete TNF-α (4), and may explain the lack of association with TNF-α if visceral adipose tissue similarly does not secrete TNF-α. In fact, a study of obese women found that TNF-α mRNA expression was significantly lower in visceral fat than in abdominal subcutaneous fat (29).
This study is subject to some limitations. The study population consisted of relatively healthy elderly adults, and the extrapolation of these findings to younger individuals or frail older adults may not be appropriate. Due to the variety of medical conditions present in men and women in the study population, there is the possibility that the visceral fat associations found in the analysis were due to a classical acute-phase response to illness or injury rather than endogenous or low-grade exogenous factors. To investigate this possibility, linear regression was conducted on a cohort subset in which the subjects with the highest decile of inflammatory marker values were excluded from the analysis. The visceral fat associations retained their significance and increased slightly in strength, indicating that the acute-phase response may obscure rather than strengthen the relationship between visceral fat and inflammation (data not shown).
This study is unique in that it investigated the relationship between inflammatory marker concentrations and multiple specific fat depots, and explored these associations by race and gender group. Across all race and gender groups, visceral fat was found to be significantly associated with IL-6 and CRP. Thigh intermuscular fat had an inconsistent but significant association with inflammation in the four population groups, and there was found a trend toward lower inflammatory marker concentration with increasing thigh subcutaneous fat.
This study was supported by National Institute on Aging contracts N01-AG-6-2101, N01-AG-6-2103, and N01-AG-6-2106. This research was supported (in part) by the Intramural Research Program of the National Institutes of Health, National Institute on Aging. The authors' responsibilities were as follows—L.E.B.: data analysis and interpretation, drafting of the manuscript; A.K.: study design, data interpretation, drafting of the manuscript, and critical revision of the manuscript; A.B.N., M.K.J., L.F., S.B.K., L.H.K., M.P., L.A.S., M.V., S.M.R., B.H.G.: critical revision of the manuscript; T.B.H.: study design, interpretation of the data, and revision of the manuscript. There is no personal or financial conflict of interest among any of the authors.