Clinical Diabetes and Nutrition Section, National Institutes of Health, 4212 N 16th Street. Room 5–41, Phoenix, AZ 85016. E-mail: email@example.com
Objective: Plasma concentrations of interleukin-6 (IL-6), a proinflammatory cytokine produced and released in part by adipose tissue, are elevated in people with obesity and type 2 diabetes. Because recent studies suggest that markers of inflammation predict the development of type 2 diabetes, we examined whether circulating plasma IL-6 concentrations were related to direct measures of insulin resistance and insulin secretory dysfunction in Pima Indians, a population with high rates of obesity and type 2 diabetes.
Research Methods and Procedures: Fasting plasma IL-6 concentrations (enzyme-linked immunosorbent assay), body composition (DXA), insulin action (M; hyperinsulinemic euglycemic clamp), and acute insulin secretory responses to glucose (25 g intravenous glucose tolerance test) were measured in 58 Pima Indians without diabetes (24 women, 34 men).
Results: Fasting plasma IL-6 concentrations were positively correlated with percentage of body fat (r = 0.26, p = 0.049) and negatively correlated with M (r = −0.28, p = 0.031), but were not related to acute insulin response (r = 0.13, p = 0.339). After adjusting for percentage of body fat, plasma IL-6 was not related to M (partial r = −0.23, p = 0.089).
Discussion: Fasting plasma IL-6 concentrations are positively related to adiposity and negatively related to insulin action in Pima Indians. The relationship between IL-6 and insulin action seems to be mediated through adiposity.
In recent years, a number of studies have indicated that several humoral markers of inflammation are elevated in people with obesity and type 2 diabetes (1, 2). Based on these and other findings, it has been proposed that long-term activation of the innate immune system may be involved in the development of insulin resistance and type 2 diabetes.
One possible explanation for elevated inflammatory markers in obesity is that adipose tissue secretes a number of proinflammatory cytokines, including tumor necrosis factor-α (TNF-α) and interleukin-6 (IL-6) (1). Although immune cells, fibroblasts, endothelial cells, and monocytes have traditionally been regarded as the major sources of circulating IL-6 (3), a recent study in which adipose tissue veins were selectively catheterized has indicated that a considerable proportion of circulating IL-6 is derived from adipose tissue (4). Circulating IL-6 levels have been reported to be elevated in obese people (3, 5) and in people with type 2 diabetes (6) and to correlate with indirect measures of adiposity and insulin resistance, such as body mass index (BMI), waist-to-hip ratio (5, 7), and fasting insulin concentrations (7). However, to our knowledge, no study has examined the relationship between circulating IL-6 levels and direct measures of adiposity, insulin action, and insulin secretion. Thus, it is unclear whether the association between insulin resistance and markers of inflammation is independent of obesity.
The aim of the present study was to examine the relationship between fasting plasma IL-6 concentrations and direct measures of adiposity, insulin action, and insulin secretion in Pima Indians without diabetes, a population that has one of the highest prevalence rates of obesity, insulin resistance, and type 2 diabetes in the world (8, 9).
Research Methods and Procedures
A total of 58 Pima Indians, 24 women and 34 men, were included in this analysis (Table 1). All were participants in an ongoing longitudinal study of the pathogenesis of type 2 diabetes, were between 20 and 50 years of age, did not have diabetes according to a 75-g oral glucose tolerance test (World Health Organization criteria), were nonsmokers at the time of the study, and were healthy according to a physical examination and routine laboratory tests. No subject had clinical or laboratory signs of acute infection, dyslipidemia, hypertension, and/or a personal history of hypertension, dyslipidemia, atherosclerotic disease, autoimmune disease, or other conditions known to be associated with altered plasma IL-6 concentrations. The Tribal Council of the Gila River Indian Community and the Institutional Review Board of the National Institute of Diabetes and Digestive and Kidney Diseases approved the protocol and all subjects provided written informed consentbefore participation.
Table 1. Physical and metabolic characteristics of the study population
Mean ± SD
M-low, insulin-stimulated glucose disposal during low-dose insulin infusion; M-high, insulin-stimulated glucose disposal during high-dose insulin infusion; EMBS, estimated metabolic body size (fat free mass + 17.7 kg).
30 ± 7
Body weight (kg)
89.9 ± 20.3
Body mass index (kg/m2)
32.5 ± 6.5
Body fat (%)
32 ± 8
Fat mass (kg)
29.7 ± 11.8
Fat-free mass (kg)
60.2 ± 12.0
41 ± 6
1.62 ± 0.15
Fasting glucose (mg/dL)
84 ± 8
Fasting insulin (μU/mL)
38 ± 16
2-hour glucose (mg/dL)
122 ± 29
M-low (mg/kg EMBS per minute)
2.9 ± 1.2
M-high (mg/kg EMBS per minute)
8.3 ± 2.1
Acute insulin response (μU/mL)
218 ± 137
Fasting plasma IL-6 (pg/mL)
2.6 ± 1.5
Subjects were admitted to the National Institutes of Health Clinical Research Unit in Phoenix, Arizona, for 8 to 10 days for testing. While in the unit, subjects were fed a weight-maintaining diet (50%, 30%, and 20% of daily calories provided as carbohydrate, fat, and protein, respectively) and abstained from strenuous exercise.
Body composition was estimated by total body DXA (DPX-L; Lunar Radiation Corp., Madison, WI) with calculations of percentage of body fat, fat mass, and fat free mass as previously described (9). Waist and thigh circumferences were measured at the umbilicus in the supine and the gluteal fold in the standing positions. The waist-to-thigh ratio was calculated as an index of body fat distribution.
A 2-hour 75-g oral glucose tolerance test was performed after a 12-hour overnight fast (9) to exclude subjects with diabetes. Baseline blood samples were drawn for determination of glucose, insulin, and IL-6 concentrations using prechilled syringes and prechilled glass tubes. Plasma glucose concentrations were determined by the glucose oxidase method (Beckman Instruments, Fullerton, CA) and plasma insulin concentrations by an automated immunoassay (Access; Beckman Instruments). IL-6 concentrations were measured by a two-antibody enzyme-linked immunosorbent assay using bioptin-strepavidin-peroxidase detection (Cytokine Core Laboratory, Baltimore, MD; normal range 1.6 to 100 pg/mL, coefficient of variation 10.6% at 1.6 pg/mL and 1.4% at 100 pg/mL).
Insulin action was assessed at physiological and supraphysiological insulin concentrations during a two-step hyperinsulinemic euglycemic glucose clamp as previously described (9). Briefly, after an overnight fast, a primed continuous intravenous insulin infusion was administered for 100 minutes at a constant rate of 40 mU/m2 of body surface area per minute (low dose, M-low), followed by a second 100-minute infusion at a rate of 400 mU/m2 per minute (high dose, M-high). These infusions achieved steady state plasma insulin concentrations of 147 ± 42 μU/mL and 2681 ± 1410 μU/mL (mean ± SD), respectively. Plasma glucose concentrations were maintained at ∼100 mg/dL with a variable infusion of a 20% glucose solution. The rate of total insulin-stimulated glucose disposal (M) was calculated for the last 40 minutes of the low-dose (M-low) and high-dose (M-high) insulin infusions. M-low was also corrected for the rate of endogenous glucose output (measured by a primed [30 μCi], continuous [0.3 μCi/min] 3-3H-glucose infusion) (9). M-values were adjusted for the steady state plasma glucose and insulin concentrations as previously described (10) and were normalized to estimated metabolic body size (estimated metabolic body size = fat free mass + 17.7 kg).
Insulin secretion was measured in response to a 25-g intravenous glucose tolerance test and the acute insulin response (AIR) calculated as the average incremental plasma insulin concentration from the third to the fifth minute after the glucose bolus (9). Because even mildly elevated glucose concentrations can secondarily affect insulin secretion, only data from the 44 subjects with normal glucose tolerance were included when analyzing the correlation between IL-6 and AIR.
Statistical analyses were performed using the software of the SAS Institute (Cary, NC). Log-transformed values of M and IL-6 were used for all statistical analyses to achieve normal distributions. The relationships between fasting IL-6 concentrations and anthropometric and metabolic variables were assessed by simple linear regression models. Partial correlation analyses and multiple linear regression models were used to examine the relationship between IL-6, percentage of body fat, and insulin action, and to assess the effect of gender and glucose tolerance on IL-6.
The anthropometric and metabolic characteristics of the study population are summarized in Table 1.
The mean fasting plasma IL-6 concentration did not differ between men and women (2.5 ± 0.2 vs. 2.7 ± 0.3, p = 0.57) or between individuals with normal and impaired glucose tolerance (2.6 ± 0.2 vs. 2.5 ± 0.5, p = 0.98), before or after adjusting for percentage of body fat.
Fasting plasma IL-6 concentrations were positively correlated with percentage of body fat (r = 0.26, p = 0.049; Figure 1A), body weight (r = 0.26, p = 0.047), BMI (r = 0.35, p = 0.008), fat mass (r = 0.32, p = 0.015), waist (r = 0.32, p = 0.014) and thigh circumference (r = 0.28, p = 0.037), but not with the waist-to-thigh ratio (r = 0.17, p = 0.189).
Fasting plasma IL-6 concentrations were negatively correlated with the rate of insulin-stimulated glucose disposal (M) as assessed during the low-dose insulin infusion (M-low) (r = −0.28, p = 0.031; Figure 1B), but not with M during high-dose insulin infusion (M-high, r = −0.04, p = 0.769), fasting plasma insulin concentrations (r = 0.17, p = 0.208), AIR (n = 44, r = 0.13, p = 0.33), or fasting (r = 0.20, p = 0.127) or 2-hour glucose concentrations (r = 0.003, p = 0.983). The relationship between fasting plasma IL-6 concentration and M-low was no longer significant after adjustment for percentage of body fat in partial correlation analyses (partial r = −0.23, p = 0.089).
In the present study, we examined the relationships between fasting plasma IL-6 concentration and direct measures of adiposity, insulin action, and insulin secretion in Pima Indians. We found that fasting plasma IL-6 concentrations were positively related to adiposity and negatively related to insulin action but were not related to insulin secretion. However, after adjustment for adiposity, there was no correlation between IL-6 and insulin action.
Our finding of a positive relationship between IL-6 concentration and percentage of body fat (Figure 1A), fat mass, and waist circumference confirms results from previous studies in other populations, which indicate that plasma IL-6 concentrations are elevated in obesity (4, 5, 7). In those studies, plasma IL-6 concentrations seem to be more closely related to BMI and percentage of body fat than in the present study in Pima Indians (4, 7). This could in part be due to differences in subject selection. Two previous studies, for instance, included subjects with sleep disorders (5) and type 2 diabetes (7), conditions known to be associated with elevated IL-6 concentrations. A weaker relationship between IL-6 and BMI has recently been reported in urban Indians living in India (11) suggesting that ethnicity may also affect the relationship between IL-6 and obesity.
Previous studies have shown that several humoral markers of inflammation such as TNF-α, C-reactive protein, and complement C3 are inversely correlated with insulin action, independently of adiposity (1, 12, 13). These findings are consistent with the hypothesis that low-grade inflammation may have a pathogenic role in the development of insulin resistance and type 2 diabetes (14). Although there is limited experimental evidence suggesting that IL-6 may be involved in the pathogenesis of decreased insulin sensitivity (15), a positive association between plasma IL-6 concentrations and fasting insulin concentrations, an index of insulin resistance, has recently been reported (7). These results must be interpreted with caution, however, given that subjects with type 2 diabetes were included in that study and that type 2 diabetes is known to be associated with elevated IL-6 concentrations (2, 6). In the present study, we show that plasma IL-6 concentrations are inversely related to the rate of insulin-stimulated glucose disposal (M-low), a direct measure of insulin action, in subjects without diabetes (Figure 1B). However, after adjustment for obesity, M-low was not related to plasma IL-6 concentrations. The lack of a statistically significant association between M-low and plasma IL-6 concentration after adjusting for obesity is most likely due to the fact that adipose tissue secretes IL-6 or other factors (such as TNF-α or complement C3) that affect insulin action.
In summary, fasting plasma IL-6 concentrations are positively related to adiposity and negatively related to insulin action in Pima Indians. The relationship between IL-6 and insulin action seems to be mediated via adiposity.
The authors thank members of the Gila River Indian Community for their participation. We also acknowledge Mr. Mike Milner, Ms. Carol Massengill, the nurses of the Clinical Research Unit as well as the staff of the metabolic kitchen for their care of the patients in the studies, and the Clinical Diabetes and Nutrition Section technical staff. No outside funding/support was provided for this study.