Atherosclerosis 185 (2006)

Transcription

1 Atherosclerosis 185 (2006) Associations of leisure time physical activity, self-rated physical fitness, and estimated aerobic fitness with serum C-reactive protein among 3803 adults Katja Borodulin a,, Tiina Laatikainen a,b, Veikko Salomaa a, Pekka Jousilahti a,c a National Public Health Institute, Department of Epidemiology and Health Promotion, Mannerheimintie 166, FIN Helsinki, Finland b Greater Green Triangle University Department of Rural Health, A Flinders and Deakin University Partnership, Warrnambool, Australia c University of Helsinki, Helsinki, Finland Received 21 January 2005; received in revised form 26 May 2005; accepted 15 June 2005 Available online 25 July 2005 Abstract Objective: Serum C-reactive protein (CRP), a marker of systemic inflammation, is a risk factor for cardiovascular disease. Obesity and physical activity are associated with CRP, though population studies are sparse. Methods and results: We assessed the cross-sectional relationship of physical activity, self-rated fitness and estimated aerobic fitness with serum CRP among a population-based random sample of 3803 Finnish adults aged years. Conditioning, commuting and non-conditioning physical activity was measured by a recall and self-rated physical fitness by a questionnaire. A non-exercise test was used to estimate aerobic fitness. The combined association of aerobic fitness and waist-to-hip ratio (WHR) on CRP was also assessed. Self-rated fitness and aerobic fitness had inverse associations with CRP (adjusted for age, WHR, smoking, anti-hypertensive drugs, aspirin, diabetes, hypercholesterolemia, hormone replacement therapy and menopausal status) in both sexes (p < 0.001). An inverse age-adjusted association of conditioning and nonconditioning physical activity with CRP was found in both sexes. After further adjustments, associations remained significant in women for conditioning and non-conditioning activity (p < and p = 0.010, respectively) and borderline significant in men (p = and p = 0.070, respectively). Commuting physical activity was inversely associated with CRP among women only (p = 0.012). Conclusions: Physical activity and fitness may have an anti-inflammatory effect, which is independent of obesity Elsevier Ireland Ltd. All rights reserved. Keywords: Inflammation; Physical activity; Exercise test; Aerobic power; Cardiovascular disease prevention; Obesity 1. Introduction Serum C-reactive protein (CRP), a marker of systemic inflammation, is a recently established risk marker for cardiovascular disease [1 3]. Physical activity and cardiorespiratory fitness prevent cardiovascular disease by both improving the risk factor profile [4] and reducing disease risk independently of other risk factors [5]. Previous population studies have shown that physical activity and cardio-respiratory fitness are inversely associated with CRP, suggesting that physical activity and fitness have an anti- Corresponding author. Tel.: ; fax: address: katja.borodulin@ktl.fi (K. Borodulin). inflammatory effect [6 9]. Previous studies have, however, concentrated on either physical activity or cardio-respiratory fitness alone, not looking at different domains of physical activity. There is little evidence whether conditioning (i.e. fitness-related), commuting or non-conditioning (i.e. household chores) physical activity is related to CRP levels. Furthermore, most previous studies investigated men only and samples with limited age range. Other conditions related to increased CRP levels include elevated blood pressure, smoking, hyperlipidemia, postmenopausal hormone use and acute infection, while decreased CRP levels are associated with the use of statins and other drugs, such as aspirin, beta blockers and ACE-inhibitors [3,10,11]. Furthermore, CRP has a strong positive correla /$ see front matter 2005 Elsevier Ireland Ltd. All rights reserved. doi: /j.atherosclerosis

2 382 K. Borodulin et al. / Atherosclerosis 185 (2006) tion with obesity, when measured as a body mass index or waist-to-hip ratio (WHR) [12]. Obesity, in turn, is inversely associated with physical activity [13] and cardio-respiratory fitness [14] and directly with cardiovascular disease [15]. It is still unclear how obesity, particularly abdominal obesity, affects the association of cardio-respiratory fitness and CRP. We studied the cross-sectional associations of different domains of leisure time physical activity, self-rated physical fitness and estimated aerobic fitness with CRP in 3803 apparently healthy Finnish men and women, and further how WHR affects the association of estimated aerobic fitness and CRP. 2. Methods 2.1. Study population The study was conducted in the spring 2002 in six geographic areas of Finland, as part of the National FIN- RISK Study. A random sample of 13,500 persons, stratified by area, sex and 10-year age group, was drawn from the national population register according to the WHO MON- ICA protocol [16]. The physical activity sub-sample included 9179 men and women aged years, from which 3216 women (70%) and 2764 men (59%) participated. Individuals with self-reported cardiovascular disease or related condition (n = 379), such as myocardial infarction, stroke, coronary artery disease and coronary angioplasty, and with acute infection (n = 296) were excluded from further analyses. After excluding also individuals with missing data on any of the variables used in the analyses (n = 1502), the final study population comprised 1713 men and 2090 women. The Ethics Committee of the National Public Health Institute approved the study protocol and each participant gave a written informed consent Data collection The study protocol followed closely the WHO MONICA protocol [16]. The participants filled out self-administered questionnaires including questions on demographic and other background factors, smoking, physical activity, current acute infection, use of statins, aspirin or anti-hypertensive drugs, history of diabetes or impaired glucose tolerance, as well as hormone replacement therapy and menopausal status. At the study site specially trained nurses measured waist and hip circumferences and assessed aerobic fitness using a nonexercise fitness test. After measurements a venous blood specimen was taken, from which CRP and total serum cholesterol were analysed. Leisure time physical activity was assessed using a validated 12-month recall, adopted from the Kuopio Ischemic Heart Disease Study [17]. Physical activity was classified into three domains: conditioning (jogging, skiing, walking, etc.), commuting (walking and cycling to and from work) and non-conditioning (gardening, snow shovelling, household chores, etc.) physical activity. The recall allowed the calculation of metabolic equivalents (MET h/week), in which the metabolic cost was based on the Kuopio Ischemic Heart Disease Risk Factor Study protocol [17] and other internationally accepted norms [18]. Conditioning and non-conditioning physical activity was split to quartiles, where the cut-off points (MET h/week) were 2.4, 8.6, 20.0 and 0.9, 4.0, 10.5, respectively, for men and 3.1, 8.8, 18.0 and 1.8, 4.8, 10.2, respectively, for women. Commuting physical activity was divided to four categories: no activity, low, middle and high activity. The cut-off points were (MET h/week) 0, 1.2, 4.5 for men and 0, 2.3, 6.9 for women. Self-rated physical fitness was assessed using a standardized five-scale question with categories very poor, fairly poor, satisfactory, fairly good and very good. The amount of individuals in very poor fitness category was extremely low (n = 44), and thus the categories of very poor and fairly poor were combined together (n = 516) in the analyses. Aerobic fitness was measured by a non-exercise test, called the Polar Fitness Test (Polar Electro Oy, Kempele, Finland), which estimates aerobic power (ml/(kg min)) based on heart rate variability and subject s age, sex, height, weight and self-assessed four-scale amount of weekly physical activity. The different validity tests [19] of the Polar Fitness Test against maximal oxygen consumption measured during a maximal treadmill test report correlation coefficients of 0.80, 0.95 and 0.93, indicating that the estimated aerobic fitness is a good measure of aerobic fitness. More detailed information on the Polar Fitness Test has been reported elsewhere [20]. Estimated aerobic fitness was divided into three categories: low, middle and high, by using age and sex-specific international norms [21]. WHR was used as a measure of abdominal obesity [22]. In multifactorial models, WHR was used as a continuous variable. The association of aerobic fitness with CRP was also analysed in sex-specific tertiles of WHR with cut-off points at 0.93 and 0.99 for men, and 0.80 and 0.85 for women. For the venous blood sample, participants were advised to fast at least 4 h prior to their visit at the study site. A high-sensitivity CRP concentration was determined using an immunoturbidometric method (Orion Diagnostica, Espoo, Finland) with the Optima analyser (Thermo Electron Corporation, Vantaa, Finland) at the laboratory of the National Public Health Institute. The lowest detection level was at 0.2 mg/l. All values below the detection level were coded as 0.1. Due to a skewed distribution of CRP values, logtransformations were used in the analyses. The values were transformed back to geometric means when presenting the data in the tables and figure. Total cholesterol was analysed using the enzymatic method (CHOD-PAP, Thermo Elektron Oy, Finland). Persons with total cholesterol 6.5 mmol/l or with self-reported statin use were defined as having hypercholesterolemia.

3 K. Borodulin et al. / Atherosclerosis 185 (2006) Statistical methods We used the Statistical Package for Social Sciences (SPSS for Windows, Chicago, IL, USA) for all analyses, including a one-way analysis of covariance to examine linear trend and a 3 3 analysis of covariance to examine interaction. Analyses were adjusted first for age (Model 1) and then for age, WHR, smoking, anti-hypertensive drug use, aspirin use, diabetes or impaired glucose tolerance and hypercholesterolemia, as well as hormone replacement therapy and menopausal status in women (Model 2). Pearson correlation coefficients were used to assess multicollinearity between estimated aerobic fitness, age and waist-to-hip ratio before including them in the adjusted models. Men and women were analysed separately. The association of estimated aerobic fitness with CRP was also analysed stratified by WHR and interactions between estimated aerobic fitness and WHR on CRP levels were tested. 3. Results Table 1 describes the characteristics of the study population by sex. Self-reported physical activity and self-rated fitness were fairly similar in men and women. Distribution of estimated aerobic fitness differed between sexes as more women (43%) than men (25%) belonged to the highest fitness group. The geometric mean values of CRP concentration were 0.38 mg/l for men and 0.46 mg/l for women. The values ranged from 0.1 mg/l (= below detection limit) to 96.5 mg/l in men and from 0.1 to 93.8 mg/l in women. Table 1 Characteristics of the study population by sex Variable Men (n = 1713) Women (n = 2090) Age (years) a 46.5 (12.3) 46.3 (12.6) Waist-to-hip ratio a 0.96 (0.07) 0.83 (0.06) C-reactive protein b 0.38 (1.0) 0.46 (1.0) Estimated aerobic fitness 38.1 (7.1) 35.1 (10.3) (ml/(kg min)) a Conditioning physical activity c 8.6 (2.4, 20.0) 8.8 (3.1, 18.0) Non-conditioning physical activity c 4.0 (0.9, 10.5) 4.8 (1.8, 10.2) Commuting physical activity c 0 (0, 0.5) 0 (0, 3.2) Current smoker (%) Hypercholesterolemia (%) Anti-hypertensive drug use (%) Aspirin use (%) Diabetes or impaired glucose tolerance (%) Self-rated physical fitness Very and fairly poor (%) Satisfactory (%) Fairly good (%) Very good (%) Estimated aerobic fitness Low (%) Middle (%) High (%) Hormone replacement therapy (%) 23.6 Pre-menopausal (%) 64.7 a Means (S.D.). Geometric mean (S.E.) in mg/l. c Median (interquartile range) in MET h/week. Table 2 CRP (geometric means) according to conditioning and non-conditioning leisure time physical activity in men and women 1st Quartile 2nd Quartile 3rd Quartile 4th Quartile p-value for trend Men Conditioning physical activity (n = 429) (n = 427) (n = 429) (n = 428) Model 1 a <0.001 Model 2 b Non-conditioning physical activity (n = 432) (n = 424) (n = 429) (n = 428) Model Model Women Conditioning physical activity (n = 523) (n = 521) (n = 524) (n = 522) Model <0.001 Model <0.001 Non-conditioning physical activity (n = 520) (n = 525) (n = 523) (n = 522) Model Model b Model 2, ANCOVA, adjusted for age, waist-to-hip ratio, smoking, anti-hypertensive drug use, aspirin use, diabetes and hypercholesterolemia. In women, adjusted also for hormone replacement therapy and menopausal status.

4 384 K. Borodulin et al. / Atherosclerosis 185 (2006) Table 3 CRP (geometric means) according to commuting leisure time physical activity No commuting activity Low Middle High p-value for trend Men (n = 1190) (n = 173) (n = 169) (n = 180) Model 1 a Model 2 b Women (n = 1203) (n = 292) (n = 293) (n = 302) Model Model b Model 2, ANCOVA, adjusted for age, waist-to-hip ratio, smoking, anti-hypertensive drug use, aspirin use, diabetes and hypercholesterolemia. In women, adjusted also for hormone replacement therapy and menopausal status. Table 4 CRP (geometric means) according to self-rated physical fitness Poor Satisfactory Fairly good Very good p-value for trend Men (n = 209) (n = 688) (n = 686) (n = 130) Model 1 a <0.001 Model 2 b <0.001 Women (n = 307) (n = 852) (n = 812) (n = 119) Model <0.001 Model <0.001 b Model 2, ANCOVA, adjusted for age, waist-to-hip ratio, smoking, anti-hypertensive drug use, aspirin use, diabetes and hypercholesterolemia. In women, adjusted also for hormone replacement therapy and menopausal status. An inverse age-adjusted association of conditioning and non-conditioning physical activity with CRP was found in both sexes (Table 2). After further adjustments for smoking, WHR, anti-hypertensive drug use, aspirin use, diabetes or impaired glucose tolerance, hypercholesterolemia, hormone replacement therapy and menopausal status in women, these associations remained statistically significant in women for both conditioning and non-conditioning physical activity (p < and p = 0.010, respectively) and borderline significant in men (p = and p = 0.070, respectively). In commuting physical activity, a statistically significant inverse association with CRP was found among women (p = 0.012) but not among men (Table 3). Self-rated physical fitness had an inverse age-adjusted association with CRP in both sexes, and the association remained statistically highly significant also after further adjustments in multivariate model (p < in both sexes, Table 4). Similarly, estimated aerobic fitness was inversely associated with CRP, which remained statistically significant after further adjustment in Model 2 (p < in both sexes, Table 5). Estimated aerobic fitness had an inverse association with CRP at all WHR levels, even though the relationship was somewhat inconsistent among women in the highest WHR tertile (Fig. 1). The geometric mean CRP among men belonging to the highest fitness category and lowest WHR tertile was 0.22 mg/l compared to 0.86 among men in the lowest fitness category and highest WHR tertile. Among women the values were 0.27 and 0.97 mg/l, respectively. There were no statistically significant interactions between estimated aerobic fitness and WHR (p = in men and p = in women). Table 5 CRP (geometric means) according to estimated aerobic fitness Low Middle High p-value for trend Men (n = 549) (n = 690) (n = 429) Model 1 a <0.001 Model 2 b <0.001 Women (n = 484) (n = 713) (n = 893) Model <0.001 Model <0.001 b Model 2, ANCOVA, adjusted for age, waist-to-hip ratio, smoking, antihypertensive drug use, aspirin use, diabetes and hypercholesterolemia. In women, adjusted also for hormone replacement therapy and menopausal status. 4. Discussion Self-rated physical fitness and estimated aerobic fitness are associated with decreased levels of systemic inflammation among healthy Finnish men and women. Furthermore, regular conditioning, commuting and non-conditioning physical activity have inverse associations with CRP in women. These associations are independent of age, obesity, smoking, anti-hypertensive drug use, aspirin use, diabetes or impaired glucose tolerance, hypercholesterolemia, hormone replacement therapy and menopausal status. Estimated physical fitness has an inverse association with CRP levels across WHR tertiles. This indicates that even the most obese benefit from good aerobic fitness. Our findings agree with previous cross-sectional and prospective studies [6 8,23], reporting lower risk for elevated

5 K. Borodulin et al. / Atherosclerosis 185 (2006) Fig. 1. CRP (geometric means) according to estimated aerobic fitness and WHR. Analysed by using 3 3 ANCOVA, adjusted for age, smoking, antihypertensive drug use, aspirin use, diabetes, hypercholesterolemia, hormone replacement therapy and menopausal status. Panel A for men (p < for fitness and WHR, respectively, p = for interaction) and Panel B for women (p < for fitness and WHR, respectively, p = for interaction). CRP among the physically active groups compared to the sedentary groups. The findings on associations of CRP with self-rated fitness or aerobic fitness are sparse and conducted in small samples. Two studies [9,24] suggest inverse associations between fitness and CRP, in concordance with our results. One recent study [25] also looked at aerobic fitness and obesity together, and found that the most obese benefited from fitness more than leaner subjects. Also in the present study, CRP values decreased steeply with increasing aerobic fitness among the most obese subjects, even though the interaction between estimated aerobic fitness and WHR was not statistically significant. We did not find any studies that had investigated different domains of physical activity in relation to CRP. Thus, our findings on conditioning, commuting and non-conditioning physical activity are among the first ones. Different types of physical activity were reported by King et al. [26], showing inverse associations of jogging, swimming, cycling, aerobic dancing and weight lifting, but not gardening, with CRP. They analysed men and women together, which makes their results not fully comparable with ours. To obtain benefits from leisure time physical activity, one has to be regularly active. For conditioning physical activity among women, we detected 27% lower adjusted CRP value in the most active quartile compared to the least active quartile. The most active quartile performed physical activities at least 18.0 MET h/week, which translates to some 3.5 h of brisk walking per week, regularly during a 12-month period. This amount meets the current recommendation of physical activity, 30 min on most days of the week [27]. Similarly, women in the highest activity group engaged in commuting activities for 1.5 h/week and more had 20% lower CRP value than women not doing any commuting activity. Commuting physical activity was not related to lower CRP levels in men. The highest cut-off point of commuting activity in men was low, corresponding to around 1 h of brisk walking per week. According to the current physical activity recommendation, 1 h of physical activity per week may not be enough to gain health benefits. All measured physical activity domains of the present study were inversely associated with CRP in a dose response manner. Recently, it has been stated that the intra-individual variation in CRP measurement is considerable and that three serial measurements should be implemented to accurately determine CRP [28]. Clearly, our study design was limited with only a single measurement of CRP. In addition, it has been shown that cardiovascular risk factors are related to both CRP and physical activity and that the association between CRP and physical activity may be confounded by these cardiovascular risk factors, such as diabetes, cholesterol, high-density lipoprotein cholesterol and blood pressure [4,29]. As importantly, drug use, such as statins, aspirin and beta blockers, ACE inhibitors may have an effect on CRP level [3,10].We, therefore, adjusted the analyses for diabetes, hypercholesterolemia, hypertension treatment and aspirin use. Unfortunately, our questionnaire on hypertension treatment did not allow us to distinguish between different drugs. Another concern is that people who suffer from long-term illnesses and who have higher CRP levels might lead sedentary life and have poor aerobic fitness. We, therefore, excluded individuals with history of cardiovascular disease or related condition and with an acute infection to overcome the problem of selection. Another limiting factor of our study, though present in most previous studies as well, was the use of self-reported data on physical activity. Naturally, some recall bias cannot be avoided. However, the detailed 12-month recall most probably separates the physically active people from the sedentary ones, as we found by using the quartiles of physical activity in our analyses. Also, a recall typically cannot assess every possible type of physical activity behaviour, allowing researchers to detect only a fraction of total physical activity. If considering all leisure time, we can assume that not all activities were recorded by our recall, and most likely non-conditioning

6 386 K. Borodulin et al. / Atherosclerosis 185 (2006) activities were the least reported. In our results, we found only a marginally significant association of CRP with conditioning and non-conditioning physical activity among men. This could result from underreporting, but also from differences in statistical power, as our dataset had fewer men than women. We used a non-exercise test to estimate aerobic fitness. A limitation of the method is that in addition to the biological measurements, data on self-rated physical activity is also used in the estimation of physical fitness. However, the test has been validated and the estimated aerobic fitness correlated well with the maximal treadmill test results [19]. It should also be born in mind that the most accurate methods like maximal treadmill tests are not feasible in population studies of this magnitude like ours. Including both physical activity and aerobic fitness in the study design can be seen as strength of the study, and have not been studied simultaneously earlier. The mechanism how physical activity and aerobic fitness affect the inflammation process remains unclear. The current knowledge indicates that certain cytokines, such as tumor necrosis factor-alpha and interleukin-6, are stimulators of CRP production from liver. Cytokines like interleukin- 6 are produced largely by adipose tissue [30], which leads to the interpretation that reducing adipose mass by exercise also reduces levels of inflammation markers. Previous studies and our results reported in this paper were adjusted for obesity, suggesting that long-term exercise may downregulate inflammation independent of obesity. This issue clearly deserves more research to understand the roles of obesity and exercise in inflammation. The physically most active and the fittest persons have the lowest levels of CRP at all levels of abdominal obesity, suggesting that leisure time physical activity and aerobic fitness have an anti-inflammatory effect. This anti-inflammatory effect might explain part of the CVD risk reduction associated with physical activity and aerobic fitness. Acknowledgements This study was partially funded by the Finnish Ministry of Education and Polar Electro Oy. References [1] Libby P, Ridker PM, Maseri A. Inflammation and atherosclerosis. Circulation 2002;105(9): [2] Jousilahti P, Salomaa V, Rasi V, Vahtera E, Palosuo T. The association of C-reactive protein, serum amyloid a and fibrinogen with prevalent coronary heart disease baseline findings of the PAIS project. Atherosclerosis 2001;156(2): [3] Pearson TA, Mensah GA, Alexander RW, et al. Markers of inflammation and cardiovascular disease: application to clinical and public health practice: a statement for healthcare professionals from the Centers for Disease Control and Prevention and the American Heart Association. Circulation 2003;107(3): [4] McMurray RG, Ainsworth BE, Harrell JS, Griggs TR, Williams OD. Is physical activity or aerobic power more influential on reducing cardiovascular disease risk factors? Med Sci Sports Exerc 1998;30(10): [5] Lakka TA, Venalainen JM, Rauramaa R, Salonen R, Tuomilehto J, Salonen JT. Relation of leisure-time physical activity and cardiorespiratory fitness to the risk of acute myocardial infarction. N Engl J Med 1994;330(22): [6] Ford ES. Does exercise reduce inflammation? Physical activity and C-reactive protein among U.S. adults. Epidemiology 2002;13(5): [7] Geffken DF, Cushman M, Burke GL, Polak JF, Sakkinen PA, Tracy RP. Association between physical activity and markers of inflammation in a healthy elderly population. Am J Epidemiol 2001;153(3): [8] Abramson JL, Vaccarino V. Relationship between physical activity and inflammation among apparently healthy middle-aged and older US adults. Arch Intern Med 2002;162(11): [9] Church TS, Barlow CE, Earnest CP, Kampert JB, Priest EL, Blair SN. Associations between cardiorespiratory fitness and C-reactive protein in men. Arterioscler Thromb Vasc Biol 2002;22(11): [10] Joynt KE, Gattis WA, Hasselblad V, et al. Effect of angiotensinconverting enzyme inhibitors, beta blockers, statins, and aspirin on C- reactive protein levels in outpatients with heart failure. Am J Cardiol 2004;93(6): [11] Di Napoli M, Papa F. Angiotensin-converting enzyme inhibitor use is associated with reduced plasma concentration of C-reactive protein in patients with first-ever ischemic stroke. Stroke 2003;34(12): [12] Visser M, Bouter LM, McQuillan GM, Wener MH, Harris TB. Elevated C-reactive protein levels in overweight and obese adults. JAMA 1999;282(22): [13] Lahti-Koski M, Pietinen P, Heliovaara M, Vartiainen E. Associations of body mass index and obesity with physical activity, food choices, alcohol intake, and smoking in the FINRISK studies. Am J Clin Nutr 2002;75(5): [14] Ross R, Katzmarzyk PT. Cardiorespiratory fitness is associated with diminished total and abdominal obesity independent of body mass index. Int J Obes Relat Metab Disord 2003;27(2): [15] 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;23(9): [16] World Health Organization. The World Health Organization MON- ICA Project (monitoring trends and determinants in cardiovascular disease): a major international collaboration WHO MONICA Project Principal Investigators. J Clin Epidemiol 1988;41(2): [17] Lakka TA, Salonen JT. Intra-person variability of various physical activity assessments in the Kuopio Ischaemic Heart Disease Risk Factor Study. Int J Epidemiol 1992;21(3): [18] Ainsworth BE, Haskell WL, Whitt MC, et al. Compendium of physical activities: an update of activity codes and MET intensities. Med Sci Sports Exerc 2000;32(9 Suppl):S [19] Kinnunen H, Vainamo K, Hautala A, Makikallio T, Tulppo M, Nissila S. Artificial neural network in predicting maximal aerobic power. Med Sci Sports Exerc 2000;32(5):S308 [abstract 1535]. [20] Borodulin K, Lakka TA, Laatikainen T, Laukkanen R, Kinnunen H, Jousilahti P. Associations of self-rated fitness and different types of leisure time physical activity with predicted aerobic fitness in 5979 Finnish adults. JPAH 2004;1(2): [21] Shvartz E, Reibold RC. Aerobic fitness norms for males and females aged 6 to 75 years: a review. Aviat Space Environ Med 1990;61(1):3 11. [22] Ashwell M, Cole TJ, Dixon AK. Obesity: new insight into the anthropometric classification of fat distribution shown by computed tomography. Br Med J (Clin Res Ed) 1985;290(6483):

7 K. Borodulin et al. / Atherosclerosis 185 (2006) [23] Wannamethee SG, Lowe GD, Whincup PH, Rumley A, Walker M, Lennon L. Physical activity and hemostatic and inflammatory variables in elderly men. Circulation 2002;105(15): [24] LaMonte MJ, Durstine JL, Yanowitz FG, et al. Cardiorespiratory fitness and C-reactive protein among a tri-ethnic sample of women. Circulation 2002;106(4): [25] Aronson D, Sheikh-Ahmad M, Avizohar O, et al. C-reactive protein is inversely related to physical fitness in middle-aged subjects. Atherosclerosis 2004;176(1): [26] King DE, Carek P, Mainous AG. 3rd,Pearson WS Inflammatory markers and exercise: differences related to exercise type. Med Sci Sports Exerc 2003;35(4): [27] U.S. Department of Health and Human Services. Physical Activity and Health: A Report of the Surgeon General. Atlanta, GA, USA: Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion; [28] Koenig W, Sund M, Frohlich M, Lowel H, Hutchinson WL, Pepys MB. Refinement of the association of serum C-reactive protein concentration and coronary heart disease risk by correction for within-subject variation over time: the MONICA Augsburg studies, 1984 and Am J Epidemiol 2003;158(4): [29] Pihl E, Zilmer K, Kullisaar T, Kairane C, Pulges A, Zilmer M. High-sensitive C-reactive protein level and oxidative stress-related status in former athletes in relation to traditional cardiovascular risk factors. Atherosclerosis 2003;171(2): [30] Mohamed-Ali V, Pinkney JH, Coppack SW. Adipose tissue as an endocrine and paracrine organ. Int J Obes Relat Metab Disord 1998;22(12):