Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Arthritis Rheum. Author manuscript; available in PMC 2011 November 1.
Published in final edited form as:
PMCID: PMC2962724

Abdominal Adiposity in Rheumatoid Arthritis: Association with Cardiometabolic Risk Factors and Disease Characteristics



Abdominal adiposity, especially visceral adiposity, is an emerging cardiometabolic risk factor. How abdominal fat is distributed in rheumatoid arthritis (RA) and its RA-related determinants have not been explored.


Men and women with RA were compared to non-RA controls from the Multi-Ethnic Study of Atherosclerosis. Participants underwent anthropometric measures and quantification of visceral and subcutaneous fat areas (VFA, SFA) using abdominal computed tomography.


A total of 131 RA patients were compared with 121 controls. Despite similar body mass index and waist circumference between the RA and control groups, the adjusted mean VFA was 45cm2 higher (+51%) for RA vs. control men (p=0.005) but not significantly different by RA status in women. The adjusted mean SFA was 119cm2 higher (+68%) for RA vs. control women (p<0.001) but not significantly different by RA status in men. Elevated VFA (>75th percentile) was associated with a significantly higher adjusted probability of having an elevated fasting glucose, hypertension, or the composite definition of the metabolic syndrome for the RA group compared with controls. Within the RA group, rheumatoid factor seropositivity and higher cumulative prednisone exposure were significantly associated with a higher mean adjusted VFA. Higher C-reactive protein levels and lower Sharp scores were significantly associated with both VFA and SFA.


The distribution of abdominal fat differs significantly by RA status. Higher VFA in men with RA, and the more potent association of VFA with cardiometabolic risk factors in men and women with RA, may contribute to cardiovascular risk in RA populations.


Body composition has recently emerged as an important determinant of health outcomes. More than a passive storage depot, adipose tissue is a dynamic and metabolically active organ with the ability to elaborate mediators with widespread effects on metabolism, immune function, and vascular homeostasis (1). In particular, adipose tissue deposited around the mesentery and omentum (visceral fat) is highly associated with insulin resistance and cardiovascular disease (CVD) (24). In contrast, subcutaneous adipose tissue (concentrated around the hips and buttocks) is less strongly associated with CVD, and may even exerta protective effect in women (5).

Rheumatoid arthritis (RA) is a highly inflammatory systemic autoimmune disorder affecting approximately 1–2% of adults, frequently resulting in significant joint deformity and disability. Total body fat is increased, and skeletal muscle is decreased, in RA patients compared to matched controls, with both inflammatory and non-inflammatory factors contributing to these differences (6, 7). Whether this increase is reflective of increased visceral fat, or fat in an adipose depot associated with less metabolic and CV risk (i.e. subcutaneous fat), is currently unknown, as no studies have heretofore reported quantification of visceral or subcutaneous fat in RA patients.

Abdominal computed tomography (CT) scanning at the L4/L5 interspace (level of the umbilicus) with quantification of visceral fat area (VFA) and subcutaneous fat area (SFA) of this single section is validated, reproducible, and has been the most frequently utilized representation of visceral and subcutaneous fat mass in epidemiologic investigations (8, 9). Cross-sectional abdominal fat area at this level is highly correlated with the total volume of visceral fat in the compartment (10).

For this cross-sectional investigation, we sought to identify the association of RA disease status with CT measures of abdominal fat, adjusting for potentially confounding sociodemographic, lifestyle, and comorbid disease characteristics. Further, we explored whether the magnitude of the association of abdominal fat measures with cardiometabolic risk factors differed between RA patients and controls. Finally, we sought to identify the RA disease-related characteristics with the strongest associations with abdominal fat measures. We hypothesized that abdominal fat would be quantitatively increased in RA relative to controls and that this excess abdominal fat load would contribute to cardiometabolic risk in RA patients.


Study Participants

RA Patients

RA patients were men and women participating in ESCAPE RA (Evaluation of Subclinical Cardiovascular disease And Predictors of Events in Rheumatoid Arthritis), a cohort study investigating the prevalence, progression, and risk factors for subclinical cardiovascular disease in RA described in detail previously (11). ESCAPE RA participants were recruited from patients followed at the Johns Hopkins Arthritis Center and by referral from local rheumatologists. Among 351 consecutive patients screened, 188 (53%) were ineligible or declined participation, resulting in 163 RA patients enrolled through the Johns Hopkins Arthritis Center. The remaining 34 enrolled participants were referred from community rheumatologists. All of the 197 enrolled patients met American College of Rheumatology 1987 classification criteria for RA (12) and were 45–84 years of age without prior pre-specified cardiovascular events, peripheral arterial disease or peripheral arterial vascular procedures, implanted pacemaker or defibrillator devices, or atrial fibrillation. Participants weighing over 300 pounds were excluded due to weight limitations of cardiovascular imaging equipment.

A two-thirds sample of the RA group (n=131) underwent abdominal CT scanning for determination of abdominal fat and muscle areas. Sequential participants underwent abdominal CT scanning concurrent with cardiac CT scanning until the two-thirds sample was met. The study was approved by the Institutional Review Board of the Johns Hopkins Hospital with all participants providing written informed consent prior to enrollment. Enrollment began in October 2004 and concluded in May 2006.

Non-RA Controls

Non-RA controls were participants in the Multi-Ethnic Study of Atherosclerosis (MESA). A description of the MESA design and methods has been published (13). Briefly, recruitment for the MESA study was conducted through mailed letters of solicitation at most sites, and random digit dialing at one site (the University of California, Los Angeles). Interested individuals were pre-screened via phone interview and enrolled through the MESA field center that solicited their interest. MESA participants from 5 sites (Wake Forest University, Columbia University, the University of Minnesota, Northwestern University, and the University of California, Los Angeles) underwent abdominal CT scanning for evaluation of aortic calcification at the second or third study visit. An unmatched convenience sample of these (n=121), randomly selected except for the exclusion of those currently using medications commonly used as disease modifying agents to treat RA, comprised the non-RA control group for these analyses. Identifying RA patients based on RA medications has been shown to be a more reliable method for identification than self-report of the disease (14). Scanning for this group occurred between April 2003 and August 2005. While VFA was available on all 121 controls, complete SFA (and thus also TFA) measures were available on only 41 controls (women n=16 and men n=25) due to a reduced scanning field-of-view. Considering the number of RA patients and the distribution of SFA in the RA group, the smallest detectable significant difference in SFA was calculated to be 99 cm2 for women and 76 cm2 for men, assuming a two-tailed alpha of 0.05 and 80% power.


Abdominal CT Scanning and Anthropometrics

For both the RA and control groups, the cross-sectional abdominal CT image from the level of the interspace between the fourth and fifth lumbar vertebrae was selected. A single trained reader (M.A.), masked to the clinical characteristics of the participants, quantified total fat area (TFA), VFA, and SFA for all of the scans using the National Institutes on Aging Musculoskeletal Analysis Program (MAP), as previously described (15). RA participants underwent abdominal CT scanning on the same CT scanner used for Baltimore MESA procedures.

Anthropometric measures (height, weight, body mass index (BMI), and waist and hip circumferences) were performed similarly in RA and control participants as previously described (15).


With the exception of data acquisition specific to the RA disease state, all data were collected using the same questionnaires and procedures for both the RA and control participants. RA study personnel were trained and certified in MES A procedures. Questionnaires were administered by study personnel after anthropometrics and phlebotomy, but just prior to CT scanning. Questionnaires were scanned into a database using commercial software (Teleform (Cardiff); Vista, California). Data checks for consistency were performed at the time of entering each record into the database, and when the completed dataset was compiled at the MESA coordinating center for both the RA and control groups.

Sociodemographic and Lifestyle Covariates

Age, gender, race/ethnicity, educational attainment, and current and past smoking were assessed from patient-self report. Physical activity was assessed with the 7-Day Physical Activity Recall Questionnaire (16) with the weekly total of physical activity for intentional exercise activities (moderate or brisk walking for exercise, and moderate or vigorous individual or team sports and conditioning activities) calculated for each participant. Duration of television watching, a measure of sedentariness, was also assessed.


Depressive symptoms were assessed using the Center for Epidemiologic Studies Depression Scale (CES-D)(17). Thyroid disease was classified based on the use of thyroid replacement therapy. Diabetes was defined as a fasting serum glucose>126 mg/dL or use of anti-diabetic medications. Hypertension was defined as a blood pressure>140/90 mmHg or use of antihypertensive medication. The so-called metabolic syndrome was defined according to the National Cholesterol Education Program – Adult Treatment Panel III (ATP III) criteria (18).

RA Disease Characteristics

Forty-four joints were examined for swelling, tenderness, deformity, and surgical replacement or fusion by a single trained assessor. RA disease duration was assessed by patient self-report from the date of diagnosis. RA disease activity was calculated using the Disease Activity Score for 28 joints with CRP (DAS28-CRP) (19). Current and past use of glucocorticoids, biologic and non-biologic disease modifying antirheumatic drugs (DMARDs) was queried by detailed examiner-administered questionnaires. The correlation of prior self-reported glucocorticoid exposure compared against medical records in a subset of 20 participants was 0.92.

The 21-item Stanford Health Assessment Questionnaire (HAQ)(20) was used to assess disability related to common activities. Single view, anterior-posterior radiographs of the hands and feet were obtained and scored using the Sharp-van der Heijde (SvH) method (21) by a single, trained radiologist blinded to patient characteristics.

Laboratory Covariates

Fasting serum and plasma samples were collected on the day of body composition analysis and stored at −80 degrees C. Assays for ESCAPE and MESA samples were processed in the same laboratory (Laboratory for Clinical Biochemistry Research, University of Vermont). High sensitivity C-reactive protein (CRP) and IL-6 were measured as previously described (22). Plasma lipids and glucose were measured by standard assays; LDL-cholesterol was estimated in plasma specimens having a triglyceride value <400 mg/dL using the Friedewald equation. Rheumatoid factor (RF) was assessed by ELISA, with seropositivity defined at or above a level of 40 units. Anti-CCP antibody was assessed by ELISA, with seropositivity defined at or above a level of 60 units.

Statistical Analysis

The distributions of all variables were examined. Means and standard deviations were calculated for all normally distributed continuous variables, while medians and interquartile ranges were calculated for continuous variables that were not normally distributed. For categorical variables, counts and percentages were calculated. Transformation of variables was used for highly skewed variables (e.g. VFA, SFA) to satisfy the requirements of regression modeling. Differences in means for normally distributed continuous variables were compared using t-tests and differences in medians for non-normally distributed continuous variables were compared using the Kruskal-Wallis test. Differences in proportions were compared using the chi-square goodness of fit test or Fisher’s exact test, as appropriate.

The associations of RA status with abdominal adiposity measures were modeled using multivariable linear regression, first in unadjusted models in which RA status was the only covariate modeled. Because of important gender differences in the outcomes and in the associations with many of the potential confounders, analyses were conducted in gender-specific groups. Next, extended models were fit including RA status and covariates for potential confounders (age, ethnicity, education, exercise, hours of television watching, depression, current and former smoking, thyroid disease, and use of hormone replacement therapy). Menopausal status was not modeled due to collinearity with age. The Shapiro-Wilk test was used to examine normality of the modeled outcome variables across the extent of independent variables in multivariable modeling. Variance inflation factors were calculated to ensure that variables with excessive collinearity were not modeled simultaneously. For each model, studentized residuals were calculated and plotted against predicted values to visually inspect equal variance assumptions.

Next, the associations of adiposity measures with cardiometabolic risk factors (i.e. the ATPIII defined metabolic syndrome and its components) were explored for the RA and control groups using linear regression, in simple models and with adjustment for the covariates listed above. Additional adjustment for the use of antihypertensive and lipid lowering medication was performed. Heterogeneity by RA status was tested using Analysis of Covariance (ANCOVA).

Finally, the associations of RA characteristics with the normally transformed abdominal adiposity measures were modeled using linear regression within the RA group. Covariate selection, model fitting, and model checking was performed similarly to the methods described above.

All statistical calculations were performed using Intercooled Stata 10(StataCorp, College Station, TX). In all tests, a two-tailed α of 0.05 was defined as the level of statistical significance.


A total of 131 RA patients and 121 controls underwent abdominal adiposity assessment. While VFA was available on all 121 controls, complete SFA (and thus also TFA) measures were available on only 41 controls (women n=16 and men n=25) due to a reduced scanning field-of-view. There were no significant differences in sociodemographic characteristics, anthropometrics, or VFA measures for controls with complete SFA measurement compared to controls without complete SFA (data not shown). Participant characteristics according to RA status are summarized in Table 1.

Table 1
Participant Characteristics According to Rheumatoid Arthritis Status

Comparison of Abdominal Adiposity Measures by RA Status

Comparisons of measures of abdominal adipose areas for the RA vs. control groups, according to gender, are depicted in Figure 1. For women, the adjusted mean TFA was 118cm2 higher for RA vs. control women (Figure 1.A.), representing a 41% difference (p=0.004). For men, the adjusted mean TFA was 45cm2 higher for RA vs. control men (Figure 1.A.), representing a 15% difference; however, this difference was not statistically significant. After adjustment, mean VFA did not differ for women by RA status (Figure 1.B.). In contrast, the adjusted mean VFA was 45cm2 higher for RA vs. control men (Figure 1.B.), representing a 51% difference (p=0.005). For women, the adjusted mean SFA was 119cm2 higher for RA vs. control women (Figure 1.C.), a 68% difference (p<0.001). In contrast, the adjusted mean SFA was not significantly different in men with RA compared to male controls (Figure 1.C.).

Figure 1
Crude and Adjusted Associations of RA Status with Abdominal Computed Tomography-Derived Adiposity Measures

Associations of Abdominal Adiposity Measures with Cardiometabolic Risk Factors by RA Status

For both the RA and control groups, higher levels of BMI, waist circumference, VFA and SFA were each associated with a higher prevalence of the metabolic syndrome and individual metabolic syndrome components, even after adjusting for demographics, duration of exercise and sedentary activities, depression, thyroid disease, use of antihypertensives and lipid lowering medications (Table 3). However, the associations of VFA with the metabolic syndrome and some of its components were dissimilar between the RA and control groups (Figure 2). While RA patients and controls in the lower three quartiles of VFA demonstrated similar associations of VFA with the probability of having an elevated fasting glucose, elevated blood pressure, or meeting the composite definition of the metabolic syndrome, RA patients withVFA≥75th percentile (≥167cm2) had a significantly higher probability of having these outcomes compared to the control group. Specifically, the adjusted probability of having an elevated fasting glucose was 58% for RA patients in the highest quartile of VFA compared to only 29% for controls in the same VFA quartile (p=0.008). Similarly, the adjusted probability of having an elevated blood pressure was 63% for RA patients in the highest quartile of VFA compared to only 44% for controls in the same VFA quartile (p=0.012), and the adjusted probability of meeting the composite definition for the metabolic syndrome was 74% for RA patients in the highest quartile of VFA compared to only 37% for controls in the same VFA quartile (p=0.004). In contrast, the adjusted probabilities of having a low HDL level or elevated triglycerides were not significantly different for the RA and control groups within each quartile of VFA (data not shown).

Figure 2
Adjusted Associations of Quartiles of Visceral Fat Area with Components of the Metabolic Syndrome: RA vs. Control
Table 3
Adjusted Odds Ratios for Cardiometabolic Risk Factors per Quartile Increase in Body Composition Measures for RA Patients Compared to Controls

Association of RA Characteristics with Abdominal Adiposity

Visceral Fat Area

The crude and adjusted associations of RA characteristics with VFA (square root transformed to normality, as required) are summarized in Table 2. After adjusting for sociodemographics, depression, exercise, television watching, smoking, and thyroid disease (Table 2, Model 3), RF seropositivity, log CRP, and cumulative prednisone exposure were significantly associated with VFA, while higher log total Sharp scores were inversely associated with VFA. These differences translated into adjusted mean VFAs (Table 2, Adjusted Mean column) that were 29cm2 higher in RF seropositive patients compared to seronegative (118 vs. 89cm2, 33% higher; p=0.003), 38 cm2 lower in patients with total Sharp scores≥116 (the 75th percentile) compared to those with scores≤21 (the 25th percentile) (90 vs. 128cm2, 30% lower; p=0.008), and 32 cm2 higher in patients with a CRP≥8.7 (the 75th percentile) compared to those with a level≤1.1 (the 25th percentile) (117 vs.85cm2,38% higher; p=0.015). For cumulative prednisone exposure, the association with VFA was not linear across exposure level. Only patients in the highest quartile of cumulative exposure (a dose of 9 total grams or higher) had significantly higher median visceral fat area compared to those in the lowest exposure group (no prior prednisone exposure) (data not shown) (i.e. 138 vs. 100cm2, 38cm2 higher in patients with a cumulative prednisone exposure of ≥9 grams compared to those with no prior exposure; p=0.010). The combined associations of RF and cumulative prednisone with VFA are shown in Figure 3. Patients with both RF and acumulative prednisone exposure ≥9 grams had an 87 cm2 higher adjusted mean VFA compared to those with neither of these characteristics (p<0.001), a more than doubling of VFA. The associations of RF and cumulative prednisone with VFA were merely additive (i.e. p-value for the interaction between RF and cumulative prednisone=0.53 in the adjusted model).

Figure 3
Adjusted Combined Associations of Rheumatoid Factor Status and Cumulative Prednisone with Visceral Fat Area
Table 2
Associations of RA Characteristics with Visceral and Subcutaneous Fat Areas

Subcutaneous Fat Area

The crude and adjusted associations of RA characteristics with SFA (square root transformed to normality, as required) are summarized in Table 2. After adjusting for sociodemographics, depression, exercise, television watching, smoking, and thyroid disease (Table 2, Model 3), CRP was significantly associated with SFA, while higher total Sharp scores were inversely associated with SFA. These differences translated into adjusted mean SFAs (Table 2, Adjusted Mean column) that were 43 cm2 lower for inpatients with total SvH scores ≥116 (the 75th percentile) compared to those with scores≤21 (the 25th percentile) (18% lower; p=0.047), and 90 cm2 higher in patients with a CRP≥8.7 (the 75th percentile) compared to those with a level≤1.1 (the 25th percentile) (31% higher; p=0.002).


These cross-sectional findings demonstrate that RA patients show striking differences in the distribution of abdominal fat, even after adjusting for important potential confounders, despite no significant differences in BMI or waist circumference when they are compared with non-affected individuals. Higher visceral fat was linked more strongly to some cardiometabolic risk factors in the RA group compared to non-RA controls. Within the RA group, a number of inflammatory and non-inflammatory factors may contribute to these observed differences, including factors that are modifiable within the context of RA disease management (e.g., limiting cumulative exposure to glucocorticoids).

This investigation, to our knowledge, is the first to report abdominal adiposity measures in RA patients. Our findings linking higher VFA to RA in men and higher SFA to RA in women may have important potential implications for their subsequent risk for CV disease. For example, in the Framingham Heart Study (23) both SFA and VFA were associated with CV risk factors (systolic and diastolic blood pressures, total and HDL cholesterol, triglycerides, and fasting glucose); however, only visceral fat remained strongly associated with CV risk factors after accounting for other anthropometric variables. Visceral fat, in particular, has been associated with downstream CVD outcomes, including aortic stiffness (24), coronary artery and abdominal aortic calcification (25), and CVD events (MI and stroke)(26). Thus, increased visceral fat in RA may account for a portion of the increased risk of CVD observed in the RA population, in which rates of CVD events and CVD mortality are increased an average of 50% compared to non-RA controls (27).

The origins of increased VFA and SFA in RA patients are likely multifactorial. In the present study, we identified two RA disease factors associated with increased VFA in both men and women: RF seropositivity and cumulative prednisone exposure. The basis of the association of RF with VFA is not clear. As there is no known biologic link between RF and adipose accumulation, the association may relate to RF as a disease severity marker. However, in our study, other markers of RA severity (i.e. anti-CCP antibodies, shared epitope alleles, etc…) were not associated with VFA. There is biologic plausibility behind the observed association of cumulative prednisone with VFA. A variety of changes in fat deposition have long been recognized to accompany endogenous and exogenous hypercortisolism (e.g. “moon face”, “buffalo hump”); however, our data are the first to explore the effect of chronic, low dose glucocorticoid therapy on abdominal fat distribution in rheumatoid arthritis. Glucocorticoids have been shown to differentially affect glucose uptake and insulin receptor signaling in visceral but not subcutaneous fat adipocyte explants (28), processes that may serve to preferentially increase free fatty acid storage within the visceral adipocyte and thus increase adipocyte size.

Many of the inflammatory cytokines that are chronically elevated in RA and are the hallmark of the disease (i.e. TNF-α, IL-6) have been shown to promote migration of mesenchymal precursors to adipose tissue depots (in a process deemed “adipotaxis”) (29), stimulate adipocyte differentiation, and reduce the sensitivity of adipocytes to signaling by insulin, leptin, and other adipocytokines. The potential end-result of these processes is an increase in visceral fat. Adipocytes and resident adipose tissue macrophages are an additional source of inflammatory cytokines. Antagonism of TNF-α was successful in reversing these processes several of these in vitro and animal studies. However, we did not detect an association between measures of current disease activity and VFA, nor did treatment with biologic DMARDs (the majority of which were antagonists of TNF-α) appear to be protective against elevated VFA. Similarly, no dynamic effects of TNF inhibition on body composition parameters were observed in a randomized clinical trial conducted in non-RA patients (30) or in the handful of small-scale studies in RA patients (3133).

Our finding that increasing visceral fat was negatively associated with articular damage seems counter-intuitive, particularly since visceral fat accumulation is associated with higher levels of systemic inflammation. However, recent studies by us (15) and others (34) have suggested that alterations in levels of the adipocytokine, adiponectin, may explain this apparent contradiction. Adiponectin activates pro-inflammatory processes in inflamed rheumatoid synovium, potentially contributing to articular damage (35). However, adiponectin expression by adipocytes is suppressed as visceral fat increases. Thus, higher abdominal adiposity may lead to lower adiponectin levels and thus less articular damage. Supporting this, an association of increasing BMI with protection against articular damage in RA has been reported (36).

A salient finding of our study is the demonstration of markedly different patterns of abdominal obesity in RA patients in the absence of differences in simple measures often used to estimate body composition (i.e. BMI, waist circumference). How this can occur is likely related to two processes 1) preferential accumulation of body fat into a specific adipose depot (i.e. the visceral compartment in men) and/or 2) equalization of the reduction in muscle by a compensatory increase in fat mass. While the present study is the first to explore adipose partitioning in RA, compensatory gain of fat as a feature of RA was suggested in studies by Roubenoff from a decade ago (37). The result of such partitioning and compensation is that markedly abnormal body composition may go unrecognized in RA patients when anthropometric measures such as BMI are used to approximate fat mass.

There are some notable limitations to our study. The exclusion of patients weighing more than 300 lbs could have introduced selection bias, as the RA patients exceeding this limit may have had proportionally more visceral and/or subcutaneous fat than the controls. However, inclusion of these individuals would likely have strengthened, not diminished, the magnitude of the associations detected. Several potential confounders not measured in the study, such as physical fitness, could have influenced the inferences about the association of RA status with body composition measures. Although self-report of physical activity was used to approximate this, this measure is subject to imprecision in measurement. Another limitation was that SFA was not available on a substantial number of control patients due to a reduced CT field-of-view. Although the controls with and without SFA measures were not substantially different in their characteristics, the loss of these patients could have reduced precision and power in comparing RA and control groups. However, differences in SFA in women were of sufficient magnitude to detect significant differences. Finally, differences in mode of recruitment and participation between the RA and non-RA groups could account for some of the differences in adiposity measures observed; however, the effect of this bias would have had to have been large to account for the large between-group differences observed in VFA and SFA measures.

In summary, we have identified striking differences in the amount and distribution of abdominal adiposity in RA patients compared to controls. Both the increase in the amount of visceral adipose in men with RA, and the larger magnitude of association for elevated levels of visceral adipose with cardiometabolic risk factors in both men and women with RA, may contribute to the recognized excess in cardiovascular risk in RA populations. Measures to reduce visceral fat (i.e. weight loss) may ameliorate cardiovascular risk to a greater extent in RA than in non-RA groups. However, identification of RA patients with increased abdominal adiposity may be challenging in the clinic, as anthropometrics were not different between the RA group and controls, despite significantly increased adiposity in the RA group. Although these data point to unmodifiable factors as contributing to the observed increase in visceral fat (i.e. RF), modifiable factors (i.e. chronic exposure to glucocorticoids) appear to have a comparable impact.


Funding Support

This work is supported by Grant Numbers AR050026-01 (JMB) and 1K23AR054112-01 (JTG) from the National Institutes of Health, National Institute of Arthritis and Musculoskeletal and Skin Diseases; a Clinical Investigator Fellowship Award from the Research and Education Foundation of the American College of Rheumatology (JTG); and the Johns Hopkins Bayview Medical Center General Clinical Research Center (Grant Number M01RR02719). Funding for this research was also made possible by the American College of Rheumatology Research and Education Foundation’s Within Our Reach: Finding a Cure for Rheumatoid Arthritis campaign. MESA is funded by contracts N01-HC-95159 through N01-HC-95166 and N01-HC-95169 from the National Heart, Lung, and Blood Institute.

We would like to thank the Johns Hopkins Bayview Medical Center General Clinical Research Center and staff for providing support for the DXA scanning used in this study and to the field center of the Baltimore MESA cohort and the MESA Coordinating Center at the University of Washington, Seattle.

We are indebted to the dedication and hard work of the ESCAPE RA Staff: Marilyn Towns, Michelle Jones, Patricia Jones, Marissa Hildebrandt, and Shawn Franckowiak, and to the participants in the ESCAPE RA study who graciously agreed to take part in this research.

Drs. Uzma Haque, Clifton Bingham III, Carol Ziminski, Jill Ratain, Ira Fine, Joyce Kopicky-Burd, David McGinnis, Andrea Marx, Howard Hauptman, Achini Perera, Peter Holt, Alan Matsumoto, Megan Clowse, Gordon Lam and others generously recommended their patients for this study.


Author Contributions

Study Design: Giles, Allison, Szklo, Bathon

Data Acquisition: Giles, Bathon, Allison

Statistical Analyses: Giles

Interpretation of Data and Manuscript Writing: All authors


1. Hotamisligil GS, Arner P, Caro JF, Atkinson RL, Spiegelman BM. Increased adipose tissue expression of tumor necrosis factor-alpha in human obesity and insulin resistance. J Clin Invest. 1995 May;95(5):2409–15. [PMC free article] [PubMed]
2. Phillips GB, Jing T, Heymsfield SB. Relationships in men of sex hormones, insulin, adiposity, and risk factors for myocardial infarction. Metabolism. 2003 Jun;52(6):784–90. [PubMed]
3. Lakka HM, Lakka TA, Tuomilehto J, Salonen JT. Abdominal obesity is associated with increased risk of acute coronary events in men. Eur Heart J. 2002 May;23(9):706–13. [PubMed]
4. Shuman WP, Morris LL, Leonetti DL, Wahl PW, Moceri VM, Moss AA, et al. Abnormal body fat distribution detected by computed tomography in diabetic men. Invest Radiol. 1986 Jun;21(6):483–7. [PubMed]
5. Tanko LB, Bagger YZ, Alexandersen P, Larsen PJ, Christiansen C. Peripheral Adiposity Exhibits an Independent Dominant Antiatherogenic Effect in Elderly Women. Circulation. 2003 Apr 01;107(12):1626–31. [PubMed]
6. Walsmith J, Roubenoff R. Cachexia in rheumatoid arthritis. Int J Cardiol. 2002 Sep;85(1):89–99. [PubMed]
7. Giles JT, Ling SM, Ferrucci L, Bartlett SJ, Andersen RE, Towns M, et al. Abnormal body composition phenotypes in older rheumatoid arthritis patients: association with disease characteristics and pharmacotherapies. Arthritis Rheum. 2008 Jun 15;59(6):807–15. [PMC free article] [PubMed]
8. Ohsuzu F, Kosuda S, Takayama E, Yanagida S, Nomi M, Kasamatsu H, et al. Imaging techniques for measuring adipose-tissue distribution in the abdomen: a comparison between computed tomography and 1.5-tesla magnetic resonance spin-echo imaging. Radiat Med. 1998 Mar-Apr;16(2):99–107. [PubMed]
9. Yoon DY, Moon JH, Kim HK, Choi CS, Chang SK, Yun EJ, et al. Comparison of low-dose CT and MR for measurement of intra-abdominal adipose tissue: a phantom and human study. Acad Radiol. 2008 Jan;15(1):62–70. [PubMed]
10. Han TS, Kelly IE, Walsh K, Greene RM, Lean ME. Relationship between volumes and areas from single transverse scans of intra-abdominal fat measured by magnetic resonance imaging. Int J Obes Relat Metab Disord. 1997 Dec;21(12):1161–6. [PubMed]
11. Giles JT, Szklo M, Post W, Petri M, Blumenthal RS, Lam G, et al. Coronary arterial calcification in rheumatoid arthritis: comparison to the multi-ethnic study of atherosclerosis. Arthritis Res Ther. 2009 Mar 10;11(2):R36. [PMC free article] [PubMed]
12. Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF, Cooper NS, et al. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum. 1988 Mar;31(3):315–24. [PubMed]
13. Bild DE, Bluemke DA, Burke GL, Detrano R, Diez Roux AV, Folsom AR, et al. Multi-ethnic study of atherosclerosis: objectives and design. Am J Epidemiol. 2002 Nov 1;156(9):871–81. [PubMed]
14. Ling SM, Fried LP, Garrett E, Hirsch R, Guralnik JM, Hochberg MC. The accuracy of self-report of physician diagnosed rheumatoid arthritis in moderately to severely disabled older women. J Rheumatol. 2000 Jun;27(6):1390–4. [PubMed]
15. Giles JT, Allison M, Bingham CO, 3rd, Scott WM, Jr, Bathon JM. Adiponectin is a mediator of the inverse association of adiposity with radiographic damage in rheumatoid arthritis. Arthritis Rheum. 2009 Sep 15;61(9):1248–56. [PMC free article] [PubMed]
16. Blair SN, Haskell WL, Ho P, Paffenbarger RS, Jr, Vranizan KM, Farquhar JW, et al. Assessment of habitual physical activity by a seven-day recall in a community survey and controlled experiments. Am J Epidemiol. 1985 Nov;122(5):794–804. [PubMed]
17. Sheehan TJ, Fifield J, Reisine S, Tennen H. The measurement structure of the Center for Epidemiologic Studies Depression Scale. J Pers Assess. 1995 Jun;64(3):507–21. [PubMed]
18. Grundy SM, Brewer HB, Jr, Cleeman JI, Smith SC, Jr, Lenfant C. for the Conference Participants. Definition of Metabolic Syndrome: Report of the National Heart, Lung, and Blood Institute/American Heart Association Conference on Scientific Issues Related to Definition. Circulation. 2004 Jan 27;109(3):433–8. [PubMed]
19. Prevoo ML, van’t Hof MA, Kuper HH, van Leeuwen MA, van de Putte LB, van Riel PL. Modified disease activity scores that include twenty-eight-joint counts. Development and validation in a prospective longitudinal study of patients with rheumatoid arthritis. Arthritis Rheum. 1995 Jan;38(1):44–8. [PubMed]
20. Wolfe F, Kleinheksel SM, Cathey MA, Hawley DJ, Spitz PW, Fries JF. The clinical value of the Stanford Health Assessment Questionnaire Functional Disability Index in patients with rheumatoid arthritis. J Rheumatol. 1988 Oct;15(10):1480–8. [PubMed]
21. van der Heijde D. How to read radiographs according to the Sharp/van der Heijde method. J Rheumatol. 2000 Jan;27(1):261–3. [PubMed]
22. Nettleton JA, Steffen LM, Mayer-Davis EJ, Jenny NS, Jiang R, Herrington DM, et al. Dietary patterns are associated with biochemical markers of inflammation and endothelial activation in the Multi-Ethnic Study of Atherosclerosis (MESA) Am J Clin Nutr. 2006 Jun;83(6):1369–79. [PMC free article] [PubMed]
23. Fox CS, Massaro JM, Hoffmann U, Pou KM, Maurovich-Horvat P, Liu CY, et al. Abdominal visceral and subcutaneous adipose tissue compartments: association with metabolic risk factors in the Framingham Heart Study. Circulation. 2007 Jul 3;116(1):39–48. [PubMed]
24. Sutton-Tyrrell K, Newman A, Simonsick EM, Havlik R, Pahor M, Lakatta E, et al. Aortic stiffness is associated with visceral adiposity in older adults enrolled in the study of health, aging, and body composition. Hypertension. 2001 Sep;38(3):429–33. [PubMed]
25. Rosito GA, Massaro JM, Hoffmann U, Ruberg FL, Mahabadi AA, Vasan RS, et al. Pericardial fat, visceral abdominal fat, cardiovascular disease risk factors, and vascular calcification in a community-based sample: the Framingham Heart Study. Circulation. 2008 Feb 5;117(5):605–13. [PubMed]
26. Mahabadi AA, Massaro JM, Rosito GA, Levy D, Murabito JM, Wolf PA, et al. Association of pericardial fat, intrathoracic fat, and visceral abdominal fat with cardiovascular disease burden: the Framingham Heart Study. Eur Heart J. 2009 Jan 9; [PMC free article] [PubMed]
27. Avina-Zubieta JA, Choi HK, Sadatsafavi M, Etminan M, Esdaile JM, Lacaille D. Risk of cardiovascular mortality in patients with rheumatoid arthritis: a meta-analysis of observational studies. Arthritis Rheum. 2008 Dec 15;59(12):1690–7. [PubMed]
28. Lundgren M, Buren J, Ruge T, Myrnas T, Eriksson JW. Glucocorticoids down-regulate glucose uptake capacity and insulin-signaling proteins in omental but not subcutaneous human adipocytes. J Clin Endocrinol Metab. 2004 Jun;89(6):2989–97. [PubMed]
29. Galvez BG, San Martin N, Rodriguez C. TNF-alpha is required for the attraction of mesenchymal precursors to white adipose tissue in Ob/ob mice. PLoS ONE. 2009;4(2):e4444. [PMC free article] [PubMed]
30. Bernstein LE, Berry J, Kim S, Canavan B, Grinspoon SK. Effects of etanercept in patients with the metabolic syndrome. Arch Intern Med. 2006 Apr 24;166(8):902–8. [PMC free article] [PubMed]
31. Marcora SM, Chester KR, Mittal G, Lemmey AB, Maddison PJ. Randomized phase 2 trial of anti-tumor necrosis factor therapy for cachexia in patients with early rheumatoid arthritis. Am J Clin Nutr. 2006 Dec;84(6):1463–72. [PubMed]
32. Metsios GS, Stavropoulos-Kalinoglou A, Douglas KM, Koutedakis Y, Nevill AM, Panoulas VF, et al. Blockade of tumour necrosis factor-alpha in rheumatoid arthritis: effects on components of rheumatoid cachexia. Rheumatology (Oxford) 2007 Dec;46(12):1824–7. [PubMed]
33. Serelis J, Kontogianni MD, Katsiougiannis S, Bletsa M, Tektonidou MG, Skopouli FN. Effect of anti-TNF treatment on body composition and serum adiponectin levels of women with rheumatoid arthritis. Clin Rheumatol. 2008 Jun;27(6):795–7. [PubMed]
34. Ebina K, Fukuhara A, Ando W, Hirao M, Koga T, Oshima K, et al. Serum adiponectin concentrations correlate with severity of rheumatoid arthritis evaluated by extent of joint destruction. Clin Rheumatol. 2009 Apr;28(4):445–51. [PubMed]
35. Tang CH, Chiu YC, Tan TW, Yang RS, Fu WM. Adiponectin enhances IL-6 production in human synovial fibroblast via an AdipoR1 receptor, AMPK, p38, and NF-kappa B pathway. J Immunol. 2007 Oct 15;179(8):5483–92. [PubMed]
36. Westhoff G, Rau R, Zink A. Radiographic joint damage in early rheumatoid arthritis is highly dependent on body mass index. Arthritis Rheum. 2007 Nov;56(11):3575–82. [PubMed]
37. Roubenoff R, Roubenoff RA, Ward LM, Holland SM, Hellmann DB. Rheumatoid cachexia: depletion of lean body mass in rheumatoid arthritis. Possible association with tumor necrosis factor. J Rheumatol. 1992 Oct;19(10):1505–10. [PubMed]