Author + information
- Received May 4, 2004
- Accepted August 9, 2004
- Published online November 16, 2004.
- Doron Aronson, MD*,* (, )
- Ron Sella, MD*,
- Muhammad Sheikh-Ahmad*,
- Arthur Kerner, MD*,
- Ophir Avizohar, MD†,
- Shmuel Rispler, MD, DSc*,
- Peter Bartha, MD‡,
- Walter Markiewicz, MD*,
- Yishai Levy, MD‡ and
- Gerald J. Brook, MD†
- ↵*Reprint requests and correspondence:
Dr. Doron Aronson, Department of Cardiology, Rambam Medical Center, P.O. Box 9602, Haifa 31096, Israel
Objectives We sought to study relationship between cardiorespiratory fitness and C-reactive protein (CRP) in subjects with the metabolic syndrome.
Background Recent studies have shown an association between the metabolic syndrome and chronic subclinical inflammation, as determined by elevated CRP. Cardiorespiratory fitness is associated with a lower risk of diabetes and improved insulin resistance.
Methods Physical fitness was assessed in 1,640 subjects using the Bruce treadmill protocol and expressed as maximal metabolic equivalents. The level of CRP was measured using a high-sensitivity assay.
Results Geometric mean CRP was calculated across quartiles of fitness after adjustment for age, gender, smoking, use of medications, and coronary disease. A strong inverse trend toward decreasing CRP levels with increasing fitness quartiles was present in subjects without metabolic abnormalities, subjects with one or two metabolic abnormalities, and subjects with the metabolic syndrome (all p ≤ 0.001). The effect of fitness was particularly robust among subjects with the metabolic syndrome. The adjusted mean CRP in subjects in the upper and lower fitness quartiles was 1.48 versus 0.93 mg/dl in subjects without metabolic abnormalities, 2.40 versus 1.66 mg/dl in subjects with one or two metabolic abnormalities, and 4.62 versus 2.20 mg/l in subjects with the metabolic syndrome (p = 0.049 for the interaction between fitness and number of metabolic abnormalities).
Conclusions Subjects with the metabolic syndrome who maintain a high fitness level have markedly lower CRP concentrations, as compared with those with a low fitness level.
The metabolic syndrome is present in one of four adults living in the U.S. The prevalence of coronary artery disease, myocardial infarction, and stroke is nearly threefold higher in subjects with the metabolic syndrome than in those without the syndrome (1). Recent studies have shown an association between the metabolic syndrome and chronic subclinical inflammation, as determined by elevated C-reactive protein (CRP) (2,3). This marker of inflammation has been shown to provide additional prognostic information on the subsequent risk of cardiovascular events in subjects with the metabolic syndrome (3).
The salutary health-related effects of cardiorespiratory fitness are mediated through the improvement of insulin resistance and its associated metabolic abnormalities, including dyslipidemia, hypertension, platelet function, fibrinolysis, and endothelial function (4). Recent studies also suggest that cardiorespiratory fitness has a favorable effect on plasma CRP levels (5). Although low cardiorespiratory fitness bears a striking clinical similarity to the metabolic syndrome, the association between fitness and the metabolic syndrome is not well characterized. We studied the relationship between fitness and CRP levels in subjects with the metabolic syndrome.
We studied 1,640 subjects who reported to the Rambam Center for Preventive Medicine for investigation of cardiovascular risk factors. The Investigational Review Committee on Human Research approved the study, and subjects signed a statement agreeing to the use of their medical information for research purposes.
Characteristics of the metabolic syndrome were based on the third report of the National Cholesterol Education Program's Adult Treatment Panel III criteria (6): 1) blood pressure ≥130/85 mm Hg; 2) triglycerides ≥1.7 mmol/l; 3) high-density lipoprotein (HDL) cholesterol ≤1.0 mmol/l for men and ≤1.3 mmol/l for women; and 4) fasting glucose ≥6.1 mmol/l. Because waist circumference was not measured in all subjects, we used a body mass index (BMI) cut point of ≥30 kg/m2for obesity, according to the World Health Organization criteria (6). Subjects with three or more criteria were diagnosed as having the metabolic syndrome.
Venous blood was collected from each subject after a 12-h fast and used for assay of glucose, total and HDL cholesterol, and triglycerides, using the Hitachi 911 AutoAnalyzer (Hitachi Inc., Tokyo, Japan) and Boehringer Mannheim (Indianapolis, Indiana)reagents. High-sensitivity CRP was measured with latex-enhanced immunonephelometry (Dade Behring, Newark, Delaware).
Assessment of cardiorespiratory fitness
Fitness was quantified using a maximal exercise test with the Bruce protocol (7). Maximal exertion was determined as achieving ≥85% age-predicted maximal heart rate and/or perceived exertion ≥17 on a 20-point Borg scale. The treadmill test time was used to predict maximal oxygen consumption (Vo2max), using a previously validated formula (7):where the weighting factor is 1 for men and 2 for women.
The correlation between predicted and measured Vo2max using this formula is excellent (r = 0.92) (7). Calculated Vo2max was expressed as maximal metabolic equivalents (METs) and used as a measure of physical fitness.
The distribution of CRP levels was highly skewed. Therefore, logarithmically transformed values of CRP (ln-CRP) were used, with results expressed as geometric mean values. Subjects were classified into quartiles according to their fitness levels. Baseline characteristics of the participating subjects, according to fitness quartile, were compared by use of analysis of variance for continuous variables and the Cochran-Armitage trend test for categorical variables.
Geometric mean values of CRP were calculated using two-way analysis of co-variance under a general linear model, with ln-CRP as the dependent variable, fitness quartile as one factor, and the severity of metabolic abnormalities as the other (0, 1, or 2, and ≥3 characteristics of the metabolic syndrome). Geometric mean values of CRP were adjusted for age, gender, smoking status, presence of coronary disease, and use of hormone replacement therapy, aspirin, and statins. Differences within the group were adjusted for multiple comparisons using the Bonferroni method.
Fitness was also tested as a continuous variable in stepwise linear regression models using ln-CRP as a dependent variable, as well as the following independent variables: fitness (METs), age, gender, smoking, presence of coronary disease, use of medications, and all components of the metabolic syndrome.
Multivariate logistic regression model was used to examine the association between the metabolic syndrome and high-risk CRP (>3.0 mg/l) (3) in relation to fitness level. All statistical analyses were performed using the SPSS statistical software (version 11.5, Chicago, Illinois).
The study population included 1,640 subjects (mean age 50 ± 10 years). The majority of subjects (65.2%) were overweight, and 340 (20.7%) had the metabolic syndrome. The clinical characteristics of the study participants, according to fitness quartiles, are presented in Table 1.Subjects in higher fitness quartiles were younger, less likely to be current smokers, and manifested a better metabolic profile, including lower BMI, lower systolic blood pressure, fasting glucose, total and low-density lipoprotein cholesterol, and higher HDL cholesterol.
The number of metabolic syndrome components, according to fitness quartiles, was 1.4 ± 1.3, 1.2 ± 1.2, 1.1 ± 1.1, and 1.0 ± 1.1 (p < 0.0001 for trend), and the percentage of subjects who were classified as having the metabolic syndrome was 32%, 27%, 24%, and 19%, respectively (p = 0.001 for trend).
There was a strong inverse linear association between fitness quartile and CRP concentrations (first quartile, 2.6 [interquartile range (IQR) 2.3 to 2.9]; second quartile, 2.1 [IQR 1.9 to 2.3]; third quartile 1.8 [IQR 1.6 to 2.0]; fourth quartile 1.4 [IQR 1.3 to 1.5] mg/l; p < 0.0001 for trend). The inverse association between fitness and CRP was present in both men (p < 0.0001) and women (p < 0.0001).
Figure 1depicts adjusted geometric mean values of CRP according to fitness quartiles and number of metabolic abnormalities. The strong inverse trend toward decreasing CRP levels with increasing fitness quartiles was present in all three groups (Bonferroni-adjusted for three comparisons: zero components, p = 0.001; one or two components, p < 0.0001; and three or more components, p < 0.0001). Similar results were obtained when patients with diabetes (n = 125) were analyzed separately (p = 0.002).
Subjects with the metabolic syndrome had higher CRP at any fitness level (Fig. 1). However, the effect of fitness was particularly apparent among subjects with the metabolic syndrome. Subjects with the metabolic syndrome who maintained a good fitness level had a markedly lower CRP concentration, as compared with subjects with a poor fitness level. There was a significant interaction between the fitness level and number of components of the metabolic syndrome in an unadjusted model containing only the main effects (p = 0.049) and a strong trend toward a significant interaction in the adjusted model (p = 0.078). The adjusted geometric mean CRP value in subjects who were in the upper fitness quartile compared with those in the lower fitness quartile was 1.48 versus 0.93 mg/l in subjects without metabolic abnormalities (p = 0.01), 2.40 versus 1.66 mg/l in subjects with one or two components of the metabolic syndrome (p = 0.01), and 4.62 versus 2.20 mg/l in subjects with the metabolic syndrome (p = 0.001).
The inverse association between fitness level and CRP remained significant when fitness was expressed as a continuous variable in a stepwise multivariate linear regression model (Table 2).For each one-unit increase in METs, the adjusted geometric mean value of CRP decreased by 0.058 mg/l (95% confidence interval [CI] 0.038 to 0.078). We also repeated the analysis using continuous rather than dichotomous variables for each component of the metabolic syndrome (BMI, HDL cholesterol, triglycerides, fasting glucose, and systolic blood pressure). In the continuous variable model, the reduction in CRP for each one-unit increase in METs was 0.042 mg/l (95% CI 0.017 to 0.062).
Figure 2shows that for each degree of metabolic abnormalities, the percentage of patients with high-risk CRP (≥3 mg/dl) was lowest among subjects with the highest fitness level and highest among subjects with the lowest fitness level.
In a multivariate logistic regression model, both fitness level (p < 0.0001) and the metabolic syndrome (p < 0.0001) were significant independent predictors of high-risk CRP. The interaction between fitness level and presence of the metabolic syndrome was statistically significant in unadjusted (p = 0.004) and adjusted (p = 0.03) models. Figure 3shows that the adjusted odds for high-risk CRP in subjects with the metabolic syndrome increased with decreasing fitness quartile.
The results of the present study show that the fitness level appears to modulate CRP concentrations independent of components of the metabolic syndrome and other factors known to influence CRP levels. In subjects with the metabolic syndrome who maintain a high fitness level, CRP levels are markedly lower than in those with a low fitness level. Furthermore, the effect of physical fitness on individuals with the metabolic syndrome is more pronounced than that in subjects without metabolic abnormalities.
There is a striking similarity between the features of the metabolic syndrome and those associated with poor fitness and physical inactivity. Improved fitness has favorable effects on serum lipids, blood pressure, glucose tolerance, platelet function, fibrinolysis, and endothelial function (4,8). Furthermore, it is well known that exercise improves insulin sensitivity (9). The results of the present study suggest that an additional salutary mechanism of improved physical fitness in relation to the risk of vascular disease is to mitigate the chronic inflammatory state in subjects with the metabolic syndrome.
An important mechanism for the chronic inflammation associated with the metabolic syndrome appears to be inflammatory cell infiltration with increased production of inflammatory cytokines in adipose tissue (10,11). In addition, metabolic abnormalities associated with insulin resistance (such as hyperglycemia and elevated free fatty acids) or insulin resistance, per se, may lead to cellular inflammatory responses. Primary insulin resistance independent of body fat mass has been shown to contribute to elevated CRP levels in those with the metabolic syndrome (12). Thus, the inverse relationship between physical fitness and CRP may reflect a fitness-induced increase in insulin sensitivity. The effect of fitness is expected to be more prominent in subjects with the metabolic syndrome, in whom a higher level of insulin resistance is present due to a genetic susceptibility.
Exercise may also affect the inflammatory response to obesity and other metabolic perturbations. Regular exercise is known to modulate immune responses in healthy individuals (13) and has been shown to attenuate inflammatory responses in other human diseases such as asthma (14). Although the mechanisms responsible for the effects of exercise on immune responses remain ill defined, recent data suggest that exercise modulates activation of the pro-inflammatory transcription factor nuclear factor-kappa B (14). In this context, an important limitation of the present study is that other markers of inflammation, such as cytokines, were not measured. In addition, our study does not directly address the relationship between changes in individual fitness level over time and inflammatory markers.
Our results have clinical implications, as they suggest that increasing physical activity and its associated improvement in fitness level should be particularly effective in mitigating the pro-inflammatory state associated with the metabolic syndrome.
Subjects with the metabolic syndrome who maintain a high fitness level have markedly lower CRP concentrations, as compared with those with a low fitness level. The effect of physical fitness on individuals with the metabolic syndrome is more pronounced than that in subjects without metabolic abnormalities, emphasizing the potential beneficial effects of exercise in subjects with clinical evidence of insulin resistance.
- Abbreviations and acronyms
- body mass index
- C-reactive protein
- high-density lipoprotein
- metabolic equivalent
- maximal oxygen consumption
- Received May 4, 2004.
- Accepted August 9, 2004.
- American College of Cardiology Foundation
- Isomaa B.,
- Almgren P.,
- Tuomi T,
- et al.
- Festa A.,
- D'Agostino R. Jr..,
- Howard G.,
- Mykkanen L.,
- Tracy R.P.,
- Haffner S.M.
- Ridker P.M.,
- Buring J.E.,
- Cook N.R.,
- Rifai N.
- Thompson P.D.,
- Buchner D.,
- Pina I.L,
- et al.
- Church T.S.,
- Barlow C.E.,
- Earnest C.P.,
- Kampert J.B.,
- Priest E.L.,
- Blair S.N.
- Grundy S.M.,
- Brewer H.B. Jr..,
- Cleeman J.I.,
- Smith S.C. Jr..,
- Lenfant C.
- McLaughlin T.,
- Abbasi F.,
- Lamendola C,
- et al.
- Pedersen B.K.,
- Hoffman-Goetz L.
- Pastva A.,
- Estell K.,
- Schoeb T.R.,
- Atkinson T.P.,
- Schwiebert L.M.