Weight Loss Reduces C-Reactive Protein Levels in Obese Postmenopausal Women
Background— C-reactive protein (CRP) has been proposed as an independent risk factor for cardiovascular disease and has been positively associated with body weight and body fatness. We examined the hypothesis that weight loss would reduce plasma CRP levels in obese postmenopausal women.
Methods and Results— In a sample of 61 obese (body mass index, 35.6±5.0 kg/m2), postmenopausal women (age, 56.4±5.2 years), we found that plasma CRP levels were positively associated with dual x-ray absorptiometry–measured total body fatness (r=0.36, P<0.005) and CT-measured intra-abdominal body fat area (r=0.30, P<0.02). Significant correlations were also found between plasma CRP and triglyceride levels (r=0.33, P<0.009) and glucose disposal measured by the hyperinsulinemic-euglycemic clamp technique (r=−0.29, P<0.03). Twenty-five of the 61 women tested at baseline completed a weight loss protocol. The average weight loss was 14.5±6.2 kg (−15.6%, P<0.0001), with losses of 10.4±5.4 kg fat mass (−25.0%, P<0.0001) and 2.8±1.4 kg fat-free mass (−6.0%, P<0.0001). Visceral and subcutaneous fat areas were reduced by −36.4% and −23.7%, respectively (P<0.0001). Plasma CRP levels were significantly reduced by weight loss: average −32.3%, from 3.06 (+0.69, −1.29) to 1.63 (+0.70, −0.75) μg/mL (P<0.0001, medians and interquartile differences). Changes in body weight and in total body fat mass were both positively associated with plasma CRP level reductions.
Conclusions— Adiposity was a significant predictor of plasma CRP in postmenopausal women on a cross-sectional basis. Moreover, caloric restriction–induced weight loss decreased plasma CRP levels. Weight loss may represent an important intervention to reduce CRP levels, which may mediate part of its cardioprotective effects in obese postmenopausal women.
Received September 10, 2001; revision received November 19, 2001; accepted November 22, 2001.
The acute phase reactant C-reactive protein (CRP) is a sensitive marker of inflammation.1 Plasma CRP levels are generally low in healthy patients without acute illness. However, through the use of appropriate high-sensitivity assays, it has been possible to investigate the relation between plasma CRP levels that were previously considered to be normal and cardiovascular disease, which has increasingly been related to inflammation.2 Such studies have demonstrated that within the low range, elevated CRP levels are independently associated with an increased risk for cardiovascular disease mortality and morbidity as well as acute coronary events in both men and women.3–7⇓⇓⇓⇓
Elevated CRP levels have been cross-sectionally associated with proxy indicators of elevated body fatness (body weight and body mass index [BMI]) and with cardiovascular disease risk factors and insulin resistance.8–11⇓⇓⇓ Although the nature of the relations among CRP, adiposity, and insulin resistance has not been clearly established, it has been proposed that adipose tissue–secreted interleukin-6 (IL-6) may mediate CRP level increases in obesity.8,9,12⇓⇓
The first aim of the present study was to examine cross-sectional relations among criterion measures of body fat, fat distribution, and CRP levels in a sample of 61 obese postmenopausal women. Second, we investigated the effects of weight loss on circulating CRP levels in a subsample of 25 women who completed a weight loss study. Our overall hypothesis was that CRP levels would be significantly related to adiposity measures and that a substantial reduction in body fatness would induce a reduction in CRP levels.
Postmenopausal obese white women in the greater Burlington, Vt, area were recruited by local advertisement. Sixty-one women were recruited and tested at baseline, whereas 25 women completed a weight loss program. Inclusion criteria were cessation of menstruation for at least 1 year, a BMI >27 kg/m2, and physical inactivity (<2 exercise periods/wk). All women were nonsmokers and nondiabetic. Other exclusion criteria were history or evidence on physical examination of cardiovascular disease, peripheral vascular disease or stroke, orthopedic limitations or history of fractures, weight loss/gain over the previous 6 months, and thyroid or pituitary disease. All subjects gave their informed consent for this study, which was approved by the Committee on Human Research and Medical Sciences of the University of Vermont. This cohort was used in previous genetic studies.13
Overview of the Protocol
Women who were eligible for the study at screening were weight-stabilized (see below) and tested during 2 inpatient visits separated by approximately 15 days at the General Clinical Research Center. Patients were weight-stable between the 2 visits. The first inpatient visit included measures of body composition, body fat distribution, and blood draws for blood lipid profile. The second inpatient visit included blood draws for the lipid profile, estradiol, CRP, and a hyperinsulinemic-euglycemic clamp. The same testing sequence was repeated after the weight loss program.
Standardization Before Testing
Before and after the weight loss protocol, participating volunteers were submitted to a weight stabilization period (within 2 kg of body weight) that lasted on average 40±21 days before pretesting (range, 22 to 125 days) and 86±43 days before posttesting (range, 22 to 162 days). Macronutrient intake was also stabilized 3 days before testing with a standard diet provided by the metabolic kitchen of the General Clinical Research Center containing 55% carbohydrate, 30% fat, and 15% protein.
Weight Loss Protocol
Women entered into a medically supervised weight loss program aimed at reducing body weight to <120% of ideal value. The program consisted of a 1200 kcal/d American Heart Association step 2 diet. Food was self-selected with dietitian supervision on macronutrient selection, with or without the use of a modified fasting supplement (Medifast, Take Shape, Jason Pharmaceuticals). Subjects were in the weight loss protocol for an average duration of 13.9±2.6 months, including weight stabilization before metabolic testing. Women were encouraged not to change physical activity habits during the weight loss protocol.
Body Composition and Fat Distribution
Body composition was determined by dual energy x-ray absorptiometry, with a Lunar DPX-L densitometer (Lunar Radiation Corp) as previously described.13 Abdominal adipose tissue areas13 were measured by CT, with a GE High Speed Advantage CT scanner (General Electric Medical Systems). This technique allows for precise measurement of the size of the adipose tissue depots located in the abdominal cavity (omental, or intra-abdominal fat) and in the abdominal subcutaneous compartment. Subjects were examined in the supine position with both arms stretched above the head. The scan was performed at the L4 to L5 vertebral level using a scout image of the body to establish the precise scanning position. Cross-sectional images were analyzed with the computer interface of the scanner. Intra-abdominal adipose tissue area was quantified by delineating the intra-abdominal cavity at the internal-most aspect of the abdominal and oblique muscle walls surrounding the cavity and the posterior aspect of the vertebral body. Adipose tissue was highlighted and computed by means of an attenuation range from −190 to −30 Hounsfield Units.
Plasma Lipid Profile, CRP, and Estradiol
Plasma triglyceride levels, total, and HDL cholesterol concentrations were measured enzymatically as described.13 Cholesterol concentrations in the HDL fraction were determined after precipitation of apolipoprotein B–containing lipoproteins with dextran sulfate.14 LDL cholesterol concentrations were calculated by means of the Friedewald equation.15 Plasma CRP levels were measured by means of a colorimetric competitive ELISA.16 The assay is based on the competition between biotinylated CRP and CRP in the sample for the coated antibody. Detection is performed with horseradish peroxidase conjugated with an avidin-biotin complex followed by the color reagent substrate orthophenylene diamine. The assay was standardized against World Health Organization standards, and the coefficient of variation was 5.14%, with the normal range from 0.18 to 5.05 μg/mL (sensitivity of 0.08 μg/mL). Plasma estradiol measurements were performed using the estradiol radioimmunoassay from Diagnostic System Laboratories Inc.
All subjects were tested after a 12-hour overnight fast and 3 days of standardized meals. An intravenous catheter was placed in an antecubital vein for 20% dextrose infusion. A second catheter was placed in the contralateral hand for blood sampling. The hand was warmed in a box (50°C to 55°C) to produce arterialized venous blood. Glucose disposal was measured by the procedure of DeFronzo et al,17 as previously described in our laboratory.18 Insulin was infused at a rate of 240 pmol/m2 per min to attain postprandial peripheral insulin levels and suppress hepatic glucose output. Blood glucose was monitored every 5 minutes, and euglycemia was maintained by infusing 20% dextrose at variable rates. The duration of the insulin infusion was such that the rate of infused glucose reached a constant value by the 2nd hour of the clamp. Total glucose disposal was calculated during the last 30 minutes of the insulin infusion as the mean rate of exogenous dextrose infusion (mg/min). Glucose disposal was also expressed relative to body size (per kilogram of fat-free mass). Glucose levels were measured by the glucose oxidase method by means of an automated analyzer (YSI Instruments). Insulin levels were determined with a double-antibody radioimmunoassay (Diagnostic Products Co). Intra-assay and interassay coefficients of variation were 4% and 10%, respectively.
Data are presented as mean±SD, unless stated otherwise. Spearman rank correlation coefficients were used to quantify the relations between plasma CRP levels and metabolic variables. Stepwise multiple regression analyses were performed to examine independent predictors of CRP levels. The model included age, body fat mass, fat-free mass, intra-abdominal adipose tissue area, triglyceride and estradiol levels, and glucose disposal expressed per kg of fat-free mass. The effects of weight loss on CRP levels and metabolic variables were tested by means of paired t tests on log10-transformed CRP values and a nonparametric Wilcoxon matched test.
Physical characteristics of the women who were tested at baseline in the study (n=61) are presented in Table 1. All women were postmenopausal and obese, with BMI values ranging from 27.4 to 52.4 kg/m2. Abdominal adipose tissue areas were also elevated, with averages of 520 cm2 and 191 cm2 for subcutaneous and visceral fat areas, respectively. Women of the study were, however, relatively healthy, as shown by their normal blood lipid profile according to accepted values.19 As expected for an obese population, plasma CRP levels were slightly higher than previously reported in nonobese populations.9,20⇓
Spearman rank correlation coefficients between plasma CRP levels and metabolic variables are shown in Table 2. Plasma CRP levels were positively associated with body weight, BMI, fat-free mass, total fat mass, and the intra-abdominal adipose tissue area. The correlation between plasma CRP levels and abdominal subcutaneous adipose tissue area was not significant. Associations between CRP levels and variables of the lipid profile were not significant, with the exception of triglyceride levels, which were positively related to CRP levels (r=0.33, P<0.009). We also found a significant negative correlation between CRP levels and glucose disposal (expressed per kg of fat-free mass). Thus, elevated CRP levels were observed primarily in women who were more obese, had more intra-abdominal fat, higher triglyceride levels, and lower insulin sensitivity. On the other hand, plasma estradiol and CRP levels were positively correlated (r=0.40, P<0.002).
In an attempt to examine independent predictors of CRP levels in the cross-sectional sample, we performed stepwise multiple regression analyses (data not shown in table form). Among the metabolic variables that were significantly associated with CRP levels, the best predictor of plasma CRP was body weight, which explained 18.1% of the variance in CRP levels (P=0.001). The other significant predictor was estradiol level (8.2% explained variance, P=0.02). In a model excluding measures of body size (body weight and fat-free mass), the best predictor of plasma CRP level was total body fat mass, which explained 12.9% variance (P=0.007).
Twenty-five of the 61 women tested at baseline completed a weight loss study (Table 3). Substantial weight loss was induced, as obese women lost, on average, 14.5±6.2 kg body weight (P<0.0001), with losses of 10.4±5.4 kg and 2.8±1.4 kg fat mass and fat-free mass, respectively (P<0.0001). We noted significant (P<0.0001) reductions in subcutaneous (115±81 cm2) and visceral fat areas (74±44cm2). The weight loss program increased HDL cholesterol concentrations (60.6%, P<0.0001) and decreased triglyceride levels (−15.0%, P<0.001) and total cholesterol/HDL cholesterol (−32.8%, P<0.0001). Glucose disposal was increased (33.2% and 41.9%, P<0.0001, for absolute and relative glucose disposal respectively). The insulin levels during the last 30 minutes of the clamp were significantly lower after weight loss (97.8±24.1 versus 87.0±14.1 μU/mL, P<0.05). However, the magnitude of the weight loss difference in the ratio of glucose disposal-to-clamp insulin levels was comparable to differences in relative glucose disposal (51.1%, P<0.0001).
Plasma CRP levels were examined in response to weight loss (Figure 1). The weight loss–induced reduction in plasma CRP levels averaged −32.3%; from 3.06 (+0.69, −1.29) to 1.63 (+0.70, −0.75) μg/mL (P<0.0001, medians and interquartile differences). The magnitude and significance of the weight loss–induced difference was similar when a paired t test was performed on log10-transformed CRP values or when a nonparametric Wilcoxon matched test was used (not shown). Correlations between changes in adiposity and changes in plasma CRP levels are shown in Figure 2. Changes in body weight and in total body fat mass were both positively associated with the reduction in plasma CRP levels. A significant correlation was found between the reduction in plasma CRP and improvements in HDL cholesterol levels.
We performed statistical analyses to examine whether hormonal status affects results of the present study. According both to clinical evaluation questionnaires and plasma estradiol levels, 5 women were taking hormone replacement for the entire trial, 18 women were not taking hormone replacement, 1 subject discontinued treatment, and 1 subject initiated it. Exclusion of hormone replacement users or of women who modified their hormone replacement status did not alter the significance and magnitude of the weight loss–induced difference in CRP levels (not shown). Although the weight loss program induced a significant reduction in plasma estradiol levels (Table 3), the correlation between changes in plasma estradiol and CRP did not reach statistical significance (r=0.33, P=0.11). Finally, 2 patients were using aspirin, either on a daily or semidaily dosage. Aspirin use remained constant throughout the study for both patients. Exclusion of these patients from the statistical analyses did not modify the effects of weight loss on plasma CRP (3.08+0.78/−1.15 versus 1.71+0.70/−0.81 μg/mL, P<0.0002, medians and interquartile ranges). Three patients were using statin treatment throughout the weight loss protocol. Exclusion of these patients from the statistical analyses did not modify the effects of weight loss on plasma CRP (3.11+0.76/−1.69 versus 1.63+0.70/−0.75 μg/mL, P<0.0002, medians and interquartile ranges).
In this study, we investigated the effects of weight loss on plasma CRP levels in obese postmenopausal women. Specifically, we tested the hypothesis that a substantial reduction in body fatness would induce a reduction in CRP levels. The physiological rationale underlying this hypothesis is that obesity has been positively associated with plasma CRP, and adipose tissue has been proposed as a factor directly modulating CRP levels.8,9⇓ We found that losses of fat mass in obese postmenopausal women were associated with proportional reductions in plasma CRP levels.
In the cross-sectional analysis of 61 obese postmenopausal women, we observed significant associations between metabolic variables and plasma CRP levels. Concordant with previous studies,9 we found positive correlations between body weight, BMI, and plasma CRP. Using more precise measures of body composition and body fat distribution, we found that body fat mass, fat-free mass, and intra-abdominal adipose tissue areas were also positively associated with circulating CRP levels. An association between plasma CRP and CT-measured visceral fat accumulation has been recently reported in another sample of men.21 Whether plasma CRP levels are related more closely to abdominal fat distribution than total body fatness per se remains unclear at the present time. The relation between CRP and waist-to-hip ratio was still significant after adjustment for BMI in one report.8 However, in the present study, we found that indicators of overall body fatness were better predictors of plasma CRP in multivariate models.
Potential mechanisms relating the degree of obesity and circulating CRP levels have not been clearly elucidated. Plasma CRP level is a sensitive marker of systemic inflammation22 that has been related to cardiovascular disease through several plausible pathways. These include the possibility that CRP levels reflect inflammation of coronary vessels related to the formation and severity of the atherogenic plaque or inflammation related to myocardial ischemia or necrosis.23 It has also been suggested that plasma CRP levels reflect the amount and activity of pro-inflammatory cytokines such as tumor necrosis factor-α, IL-1, and IL-6, which are implicated in the process of atherosclerotic plaque formation and acute coronary syndromes.8,9,12,23⇓⇓⇓ In this regard, IL-6, which is induced by both tumor necrosis factor-α and IL-1, has been proposed to play a central role in the relationship between CRP and cardiovascular disease.12 IL-6 is secreted in several sites including activated macrophages and lymphocytes but also in adipose tissue. The contribution of adipose tissue in IL-6 secretion has been proposed to be the link between plasma CRP and adiposity, as CRP synthesis in the liver is largely under the control of IL-6.12 Thus, it is possible that this mechanism explains the higher CRP levels in obese patients and the reductions observed with weight loss in the present study.
A recent study by Bastard et al24 found that weight loss induced a significant reduction in plasma and adipose tissue IL-6 levels in obese women. However, in the same study, a trend for a reduction in plasma CRP was noted but did not reach statistical significance. Lack of statistical significance in CRP levels after weight loss in the study by Bastard et al24 may have been attributable to the rather short duration of the weight loss and smaller amounts of weight lost (average of 3 kg fat mass) compared with that of the present study. Although no difference in plasma CRP was noted, these results are concordant with the hypothesis involving adipose tissue IL-6 in the modulation of hepatic CRP production. During the revision process of this report, a study was published Heilbronn et al25demonstrating a similar reduction of CRP levels in response to weight loss in women. Our study is concordant with these results.
Alternative hypotheses have been proposed with respect to the mechanisms underlying the associations that we found in the present study between insulin sensitivity, adiposity, and plasma CRP. Specifically, it has been proposed that a low-level chronic inflammatory state may induce insulin resistance and endothelial dysfunction, which would link the latter phenomena with obesity and cardiovascular disease.9 Accordingly, a recent study by Pradhan et al26 demonstrated that plasma CRP, and to a lesser extent IL-6, were both significant predictors of type 2 diabetes development. This association remained significant after adjustment for the degree of obesity. Finally, although the relation between changes in plasma estradiol and changes in CRP was not significant, it cannot be excluded that reductions in estrogen levels may have contributed to the weight loss effect on CRP. The extent to which these respective mechanisms are operative in explaining the relations among obesity, cardiovascular disease risk factors, insulin resistance, and plasma CRP remains to be established.
This study was supported in part by National Institutes of Health grants R01-DK52752 to Dr Poehlman, RR-00109, as well as a Fellowship from the American Heart Association (1999–2000) and a Fonde de la Recherche en Santé du Québec Scholarship to Dr Tchernof. We thank the staff of the GCRC (RR-109) at the University of Vermont for their cooperation. We also acknowledge the work of Denise DeFalco-McGeein RN, NPC.
- ↵Koenig W, Sund M, Fröhlich M, et al. C-reactive protein, a sensitive marker of inflammation, predicts future risk of coronary heart disease in initially healthy middle-aged men: results from the MONICA (Monitoring Trends and Determinants in Cardiovascular Disease) Augsburg Cohort Study, 1984 to 1992. Circulation. 1999; 99: 237–242.
- ↵Ridker PM, Glynn RJ, Hennekens CH. C-reactive protein adds to the predictive value of total and HDL cholesterol in determining risk of first myocardial infarction. Circulation. 1998; 97: 2007–2011.
- ↵Yudkin JS, Stehouwer CDA, Emeis JJ, et al. C-reactive protein in healthy subjects: associations with obesity, insulin resistance, and endothelial dysfunction: a potential role for cytokines originating from adipose tissue? Arterioscler Thromb Vasc Biol. 1999; 19: 972–978.
- ↵Ford ES. Body mass index, diabetes, and C-reactive protein among U.S. adults. Diabetes Care. 1999; 22: 1971–1977.
- ↵Hak AE, Stehouwer CDA, Bots ML, et al. Associations of C-reactive protein with measures of obesity, insulin resistance, and subclinical atherosclerosis in healthy, middle-aged women. Arterioscler Thromb Vasc Biol. 1999; 19: 1986–1991.
- ↵Tchernof A, Starling RD, Turner A, et al. Impaired capacity to lose visceral adipose tissue during weight reduction in obese postmenopausal women with the Trp64Arg β3-adrenoceptor gene variant. Diabetes. 2000; 49: 1709–1713.
- ↵Finley PR, Shifman RB, Williams RJ, et al. Cholesterol in high-density lipoproteins: use of Mg2+ dextran sulfate in its enzymatic measurement. Clin Chem. 1978; 24: 931–933.
- ↵Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma without the use of the preparative ultracentrifuge. Clin Chem. 1972; 18: 499–502.
- ↵Macy E, Hayes TE, Tracy RP. Variability in the measurement of C-reactive protein in healthy subjects: implications for reference intervals and epidemiological applications. Clin Chem. 1997; 43: 52–58.
- ↵DeFronzo RA, Tobin JD, Andres R. Glucose clamp technique: a method for quantifying insulin secretion and resistance. Am J Physiol. 1979; 237: E214–E223.
- ↵Dvorak R, DeNino WF, Ades PA, et al. Phenotypic characteristics associated with insulin resistance in metabolically obese but normal-weight young women. Diabetes. 1999; 48: 2210–2214.
- ↵Expert Panel on Detection, Evaluation and Treatment of High Blood Cholesterol in Adults. Summary of the second report of the National Cholesterol Education Program (NCEP) expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (adult treatment panel II). JAMA. 1993; 269: 3015–3023.
- ↵Tracy RP, Psaty BM, Macy E, et al. Lifetime smoking exposure affects the association of C-reactive protein with cardiovascular disease risk factors and subclinical disease in healthy elderly subjects. Arterioscler Thromb Vasc Biol. 1997; 17: 2167–2176.
- ↵Lemieux I, Pascot A, Prud’homme D, et al. Elevated C-Reactive protein: another component of the atherothrombotic profile of abdominal obesity. Circulation. 2001; 21: 961–967.
- ↵Lagrand WK, Visser CA, Hermens WT, et al. C-reactive protein as a cardiovascular risk factor. More than an epiphenomenon? Circulation. 1999; 100: 96–102.
- ↵Heilbronn LK, Noakes M, Clifton PM. Energy restriction and weight loss on very-low-fat diets reduce C-reactive protein concentrations in obese, healthy women. Arterioscler Thromb Vasc Biol. 2001; 21: 968–970.