Higher levels of physical activity or physical fitness are associated with lower risk of developing CHD (16,34), but the mechanisms mediating this relation are incompletely understood. Physical activity remains associated with lower CHD incidence, even after accounting for the beneficial effects of physical activity on traditional risk factors such as blood pressure, cholesterol, and glucose tolerance (20,31). Another mechanism that may play a role is inflammation. An epidemiologic study that estimated the contributions of traditional and novel cardiovascular biomarkers suggested that differences in levels of inflammatory markers may possibly explain about one-third of the inverse association between physical activity and CHD risk (24).
The evidence indicating that physical activity is associated with decreased inflammation comes primarily from cross-sectional studies reporting that physically active (or fit) persons have lower levels of inflammatory markers (primary C-reactive protein, CRP) assessed at a single time point (13). The interpretation of these data is limited, however, and it remains unclear whether physical activity reduces inflammation or whether a reduced inflammatory state serves as a marker for healthier persons who happen to exercise more. Data from several randomized controlled trials that have examined exercise interventions in relation to inflammatory markers generally have not supported a role of physical activity in decreasing inflammation (3,5–7,14,30). However, the trials have typically included small samples or were conducted in selected patient groups, such as overweight individuals (6,21,30) or patients with rheumatoid arthritis (4) or breast cancer (8). In addition, there are few data specific to different race/ethnic groups. In a small cross-sectional study of 135 women, higher levels of cardiorespiratory fitness were associated with lower CRP levels in white and American Indian, but not black, women (17).
We therefore examined the associations of physical activity with inflammatory markers, assessed at two time points, in a large multiethnic cohort of postmenopausal women.
Subjects for the present study were selected from the Women’s Health Initiative Observational Study (WHI-OS) of 93,676 diverse postmenopausal women age 50–79 yr in the observational component of the WHI (26,32). Recruitment into the WHI-OS occurred from 1994 to 1998 at 40 WHI clinical centers across the US, 10 of which were designated to recruit primarily minority participants. Women were excluded from the WHI-OS if they had serious illness with predicted survival of <3 yr; had a history of alcoholism, other drug dependency, or mental illness, including severe depression or dementia; or were considered poor candidates for long-term retention. Women were followed up with annual questionnaires collecting information on sociodemographic characteristics, lifestyle and health behaviors, and medical history. At baseline and in year 3, women were invited to the clinical centers for clinic measurements, where they were also asked to provide blood samples. Almost all women provided blood at baseline, whereas 83% of the original WHI-OS cohort did so at year 3.
To be eligible for this study, WHI-OS women had to provide blood samples and self-reported information on physical activity at both baseline and year 3, and be free of cardiovascular disease (myocardial infarction, stroke, coronary revascularization, pulmonary embolism, deep vein thrombosis, peripheral vascular disease) and cancer through year 3 because these major diseases may influence physical activity levels. We randomly selected 1355 women, stratified by race/ethnic group, with oversampling of minority groups, to obtain a balanced distribution of 301 white, 300 black, 300 Hispanic, 300 Asian/Pacific Islander (hereafter called Asian), and 154 American Indian (the maximum number satisfying eligibility criteria) women.
This study is approved by the Partners Human Research Committee of Brigham and Women’s Hospital, and all subjects signed an informed consent form on enrollment into the WHI-OS.
Assessment of physical activity
Women provided self-reported information on physical activity at baseline and year 3 by questionnaire (23). Recreational physical activity was assessed through a series of questions on the frequency and duration of walking, and of “strenuous,” “moderate,” and “mild” physical activities. Usual walking pace was also ascertained using key words and pace description (casual strolling or walking (<2 mph), average or normal (2–3 mph), fairly fast (3–4 mph), and very fast (>4 mph)). On the basis of the energy costs of these different activities (2), the weekly energy expenditure (MET·h·wk−1) was estimated for total recreational physical activity. The reproducibility of self-reports was assessed in a random sample of 569 women with two physical activity assessments 3 months apart (23). For total recreational physical activity, the intraclass correlation was 0.76, with no meaningful differences by race/ethnic groups. Previous analyses of physical activity in the WHI-OS also have indicated inverse associations of expected magnitude with CHD (19) and breast cancer risk (22), supporting the validity of the WHI-OS physical activity assessment.
Assessment of potential confounders
Potential confounders of the association between physical activity and inflammatory markers were assessed by self-report on questionnaires at baseline. These included age, education, family income, smoking, age at menopause, hormone therapy, alcohol consumption, and other dietary factors. Weight and height, used to calculate body mass index (BMI), were measured at clinical centers using standardized procedures at baseline and in year 3.
Measurement of inflammatory markers
Fasting blood samples were collected at baseline and year 3 in EDTA tubes, and after centrifugation, plasma samples were frozen and stored at −70°C. Women were asked to refrain from exercise on the morning of their blood draw. All biomarkers were assayed at a single laboratory for this study. The inflammatory markers analyzed were high-sensitivity CRP, interleukin 6 (IL-6), tumor necrosis factor α (TNF-α), leptin, adiponectin, and resistin. High-sensitivity CRP was measured using a validated high-sensitivity assay (Denka Seiken, Tokyo, Japan). IL-6 and leptin levels were determined by ultrasensitive ELISA (R&D Systems, Minneapolis, MN). TNF-α was assessed using its cell surface receptor 1, measured by an ELISA (R&D Systems). Total plasma adiponectin concentration was assessed using an ELISA method (ALPCO Diagnostics, Inc., Salem, NH). Plasma resistin level was measured using an ELISA (R&D Systems). Samples were batched so that baseline and year 3 samples were assayed together. The mean intra-assay coefficients of variation from blinded, replicate, quality control samples in this study for CRP, IL-6, TNF-α, leptin, adiponectin, and resistin were 14.3%, 20.1%, 11.9%, 9.3%, 16.3%, and 18.2%, respectively.
Participants were categorized into tertiles of weekly energy expenditure (Table 1) for analyses. For each inflammatory marker, women with values that were >3 SD from the mean were excluded. Because the values of each inflammatory marker were skewed, they were converted to the natural logarithmic values to improve normal distribution.
Characteristics of women across physical activity groups were examined first, after which the associations of physical activity with each inflammatory marker were evaluated. All analyses were conducted for the overall cohort of 1355 women and separately for each of the five race/ethnic groups. In cross-sectional analyses using measures from a single time point, linear regression was used to examine the mean values for each inflammatory marker, by physical activity group, at baseline and in year 3, adjusting for age, education, family income, BMI, smoking, alcohol consumption, age at menopause, hormone therapy, red meat intake, and fruit and vegetable intake. Tests for a linear trend across physical activity tertiles were conducted by assigning an ordinal score to the physical activity categories. These analyses were then repeated, stratifying by baseline BMI: normal weight (BMI < 25 kg·m−2), overweight (BMI = 25 to <30 kg·m−2), and obese (BMI ≥ 30 kg·m−2).
In analyses using measures from two time points, linear regression was used to examine the mean changes in inflammatory markers by change in physical activity level between baseline and year 3, initially classified as tertiles at baseline and year 3 and adjusting for the same variables as before, plus the baseline inflammatory marker level. These analyses were repeated, stratifying by baseline BMI (normal weight, overweight, obese) or by change in weight between baseline and year 3 (stable weight [±3% change between baseline and year 3 (29)], lost weight [>3% change], gained weight [>3% change]). For more clinically relevant results, physical activity was dichotomized as not meeting or meeting current recommendations (≥7.5 MET·h·wk−1) (10,12,35) in these analyses.
Table 1 shows the characteristics of women according to tertiles of recreational physical activity (3.75 MET·h·wk−1 is equivalent to brisk walking for 1 h·wk−1). Compared with less active women, more active women were younger, leaner, better educated, had a higher family income, were more likely to use postmenopausal hormones, smoked less, were more likely to consume alcohol in moderation, and ate less meat but more fruits/vegetables. They also had a better cardiovascular risk profile concerning high-density cholesterol and fasting blood glucose. White and Asian women were more physically active than Black, Hispanic, or American Indian women were.
Table 2 presents cross-sectional associations between physical activity level and the different inflammatory markers for all women, and each race/ethnic group, at baseline and in year 3. Black and Hispanic women had the highest levels of inflammatory markers; Asian women, the lowest. The levels of inflammatory markers at baseline and year 3 were highly correlated (Spearman r = 0.67–0.87). Among all women, there were strong inverse associations between physical activity and levels of inflammatory markers, whereas adiponectin showed an expected direct association. When examined as separate groups, these associations were observed more consistently among white, Hispanic, and American Indian women and less consistently in black and Asian women.
At baseline, the prevalence of obesity was highest in black women (42.6%) followed by American Indian (39.6%), Hispanic (26.7%), white (18.4%), and Asian women (7.4%). Cross-sectional associations between physical activity level and inflammatory markers at baseline, stratified by baseline BMI, were examined. Figure 1 shows the results for CRP and IL-6, chosen as exemplar markers. Higher BMI levels were associated with more inflammation, and within each BMI category, physical activity was inversely associated with each inflammatory marker level. This was generally observed for each of the different race/ethnic groups, examined separately (data not shown). In addition, within each BMI category, Asian women possessed levels of inflammatory markers that were lower than for other race/ethnic groups (data not shown).
Next, changes in inflammatory markers in relation to changes in physical activity between baseline and year 3 were investigated. There were no associations between change in physical activity level and change in inflammatory markers (data not shown). These analyses were repeated, stratifying by baseline BMI. Figure 2 shows the results for CRP and IL-6 for all women. There also was no association in analyses stratified according to BMI level. When the five race/ethnic groups were examined separately, there were no within-group associations (data not shown).
Changes in inflammatory markers between baseline and year 3, in relation to changes in physical activity, were then examined separately for women who remained weight stable, lost weight, or gained weight between baseline and year 3. American Indian (38.5%) and black women (37.0%) were most likely to gain weight during the 3-yr period, whereas the other groups showed similar proportions gaining weight: Asian = 30.2%, white = 29.3%, and Hispanic = 28.5%. As with the results in Figure 2 stratified by baseline BMI, little association was observed in these analyses stratified by change in body weight in all women (Fig. 3) and in the different race/ethnic groups (data not shown).
Finally, the analyses of Figures 1–3 were restricted to women not using hormone therapy at baseline because postmenopausal hormones influence CRP levels. The results were similar to those conducted among all women regardless of hormone use (data not shown).
In this multiethnic sample of women from the Women’s Health Initiative Observational Study, baseline and 3-year levels of inflammatory markers tended to be highest among black and Hispanic women and lowest in Asian women, reflecting in part race/ethnic differences in BMI. When using measures from a single time point, physical activity was associated with less inflammation cross-sectionally, as indicated by several inflammatory markers. These associations were more consistently observed among white, Hispanic, and American Indian women than among black and Asian women. However, there was no relation between changes in physical activity and changes in inflammatory markers between two time points for any race/ethnic group.
The discrepancy between the findings when using measures from a single time point and when examining change between two time points suggest at least two possible explanations. First, the inverse relation between physical activity and inflammation reported in many previous cross-sectional studies using measures from a single time point (13,14) may be spurious owing to other explanations for a lowered inflammatory state, apart from physical activity. Lower levels of adiposity are a candidate cause since physical activity is associated with lower body weight (11,27,36). However, in the present study, physical activity continued to show inverse associations with inflammatory markers after adjustment for, or stratification by, BMI in the cross-sectional analyses. Another plausible scenario is that, in these cross-sectional studies, a reduced state of inflammation may reflect better health status among those more able to exercise (i.e., suggesting “reverse causation”: physical activity does not cause decreased inflammation, but lower inflammatory status reflects healthier individuals who can and do exercise more). We excluded women with major diseases (cardiovascular disease and cancer), but we were unable to identify all women with proinflammatory conditions at baseline or who developed them during follow-up. We also were unable to adjust for medications (e.g., nonsteroidal anti-inflammatory drugs) that may influence inflammation.
A second explanation for the discrepant findings may be that physical activity indeed does cause decreased inflammation, but our assessment of physical activity was imprecise and unable to detect changes in physical activity. The physical activity assessment tool in this study has been tested for reproducibility and performs well, showing no meaningful differences by race/ethnic groups (23). Previous analyses of data from the WHI-OS have shown expected, inverse associations between baseline physical activity and CHD (19) and breast cancer incidence (22), indicating the validity of a single physical activity assessment. However, it is unclear how well this questionnaire can detect changes in physical activity.
A unique contribution of the present study, in addition to examining changes between two time points, is the composition of participants from five race/ethnic groups. Whereas some studies have included nationally representative samples of participants from the National Health and Nutrition Examination Survey (1,9,15), race/ethnicity-specific data on the association of physical activity and inflammation are sparse. A small cross-sectional study, the Cross-cultural Activity Participation Study, reported inverse associations between cardiorespiratory fitness and CRP levels for white and American Indian, but not black, women (17). Their findings are similar to the present results in that we observed inverse cross-sectional associations between physical activity and various inflammatory markers that were more consistent for white and American Indian women, compared with black women. Both the present study and Cross-cultural Activity Participation Study also observed higher levels of inflammatory markers in black women, compared with women from other race/ethnic groups, in part reflecting the higher BMI of black women. However, in another cross-sectional study of 3091 white, black, Hispanic, and Chinese women from the Multi-Ethnic Study of Atherosclerosis, there was no relation between physical activity and CRP levels (18).
The findings from this study, showing no association between change in physical activity and change in inflammation, are congruent with the findings from several randomized clinical trials testing exercise interventions, primarily in selected patient groups, including patients with rheumatoid arthritis (4), breast cancer (8), and intermittent claudication (33), or at high risk for heart disease (28), as well as persons with elevated CRP levels (7) or overweight and obese persons (21,30). However, two small trials of 319 young women (3) and 115 postmenopausal women (6) reported decreases in CRP levels with aerobic exercise, primarily occurring among those who were obese. A meta-analysis including 323 adults in aerobic exercise interventions calculated that the exercise intervention resulted in a nonsignificant 3% reduction in CRP levels (14). A recently published study reported that despite almost 100% adherence, an aerobic exercise intervention for 4 months among 162 sedentary men and women with high CRP levels at baseline did not reduce CRP levels (mean change in the intervention group = −0.4 mg·L−1 (95% confidence interval = −0.5 to 1.2 mg·L−1), control group = 0.5 mg·L−1 (95% confidence interval = −0.44 to 1.3 mg·L−1)) (7). However, within the intervention group, there was a significant direct relation between change in body weight or body fat and change in CRP, indicating the importance of adiposity, rather than physical activity, in determining CRP levels.
Another strength of the present study was the assessment of several inflammatory markers, not just CRP, in contrast to most previous studies of physical activity and inflammation (13). Also, the sample size was large enough to afford sufficient statistical power: a priori power calculations indicated that a sample of 300 women would yield 80% power to detect at least a 26% difference in 3-yr change in CRP, comparing extreme tertiles of change in physical activity and assuming a tracking correlation coefficient of 0.60 for the inflammatory marker. On the basis of cross-sectional studies, this difference in CRP levels between extreme categories of physical activity is reasonable (13), and levels of inflammatory markers in the present study tracked better than a correlation of 0.60 between baseline and year 3 (r = 0.67–0.87 for the various markers). Although there were only 154 American Indian women, they were included for descriptive purposes.
Although we examined several inflammatory markers, we did not adjust P values to account for multiple testing because the markers are correlated; thus, using standard corrections for multiple testing will be overly conservative. In addition, there is not universal agreement that P values should be adjusted for multiple testing (25).
In conclusion, among middle-age and older women overall, there were strong inverse associations between physical activity level and inflammatory markers assessed at a single time point. However, changes in physical activity were unrelated to changes in inflammatory markers. These data suggest a noncausal association between physical activity and inflammatory markers.
This study was supported by research grant N01 WH74311 from the National Heart, Lung, and Blood Institute (NHLBI)/National Institutes of Health. The WHI program is funded by the NHLBI, National Institutes of Health, US Department of Health and Human Services through contracts N01 WH22110, 24152, 32100-2, 32105-6, 32108-9, 32111-13, 32115, 32118-32119, 32122, 42107-26, 42129-32, and 44221. The authors thank the dedicated efforts of investigators and staff at the WHI clinical centers and coordinating center and extraordinary commitment of WHI participants.
Program Office: Jacques Rossouw, Shari Ludlam, Joan McGowan, Leslie Ford, and Nancy Geller (NHLBI, Bethesda, MD).
Clinical Coordinating Center: Garnet Anderson, Ross Prentice, Andrea LaCroix, and Charles Kooperberg (Fred Hutchinson Cancer Research Center, Seattle, WA).
Investigators and Academic Centers: JoAnn E. Manson (Brigham and Women’s Hospital, Harvard Medical School, Boston, MA), Barbara V. Howard (MedStar Health Research Institute/Howard University, Washington, DC), Marcia L. Stefanick (Stanford Prevention Research Center, Stanford, CA), Rebecca Jackson (The Ohio State University, Columbus, OH), Cynthia A. Thomson (University of Arizona, Tucson/Phoenix, AZ), Jean Wactawski-Wende (University at Buffalo, Buffalo, NY), Marian Limacher (University of Florida, Gainesville/Jacksonville, FL), Robert Wallace (University of Iowa, Iowa City/Davenport, IA), Lewis Kuller (University of Pittsburgh, Pittsburgh, PA), and Sally Shumaker (Wake Forest University School of Medicine, Winston-Salem, NC).
Women’s Health Initiative Memory Study: Sally Shumaker (Wake Forest University School of Medicine, Winston-Salem, NC).
Dr. Lee has served as a research consultant for Virgin HealthMiles. Dr. Ridker is listed as a coinventor on patents held by the Brigham and Women’s Hospital that relate to the use of inflammatory biomarkers in cardiovascular disease that have been licensed to Siemens and AstraZeneca, and has served as a research consultant to Merck, Isis, Vascular Biogenics, Amylin, and Genzyme. Dr. Manson is listed as a coinventor on a pending patent held by Brigham and Women’s Hospital that relates to inflammatory biomarkers in diabetes prediction. The other authors have reported that they have no relationships to disclose.
The results of the present study do not constitute endorsement by the American College of Sports Medicine.
1. Abramson JL, Vaccarino V. Relationship between physical activity and inflammation among apparently healthy middle-aged and older US adults. Arch Intern Med. 2002; 162: 1286–92.
2. 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): S498–516.
3. Arikawa AY, Thomas W, Schmitz KH, Kurzer MS. Sixteen weeks of exercise
reduces C-reactive protein
levels in young women. Med Sci Sports Exerc. 2011; 43 (6): 1002–9.
4. Baslund B, Lyngberg K, Andersen V, et al.. Effect of 8 wk of bicycle training on the immune system of patients with rheumatoid arthritis. J Appl Physiol. 1993; 75: 1691–5.
5. Beavers KM, Hsu FC, Isom S, et al.. Long-term physical activity and inflammatory biomarkers in older adults. Med Sci Sports Exerc. 2010; 42 (12): 2189–96.
6. Campbell PT, Campbell KL, Wener MH, et al.. A yearlong exercise
intervention decreases CRP among obese postmenopausal women. Med Sci Sports Exerc. 2009; 41 (8): 1533–9.
7. Church TS, Earnest CP, Thompson AM, et al.. Exercise
without weight loss does not reduce C-reactive protein
: the INFLAME study. Med Sci Sports Exerc. 2010; 42 (4): 708–16.
8. Fairey AS, Courneya KS, Field CJ, et al.. Effect of exercise
training on C-reactive protein
in postmenopausal breast cancer survivors: a randomized controlled trial. Brain Behav Immun. 2005; 19: 381–8.
9. Ford ES. Does exercise
reduce inflammation? Physical activity and C-reactive protein
among U.S. adults. Epidemiology
. 2002; 13: 561–8.
10. Garber CE, Blissmer B, Deschenes MR, et al.. Quantity and quality of exercise
for developing and maintaining cardiorespiratory, musculoskeletal, and neuromotor fitness in apparently healthy adults: guidance for prescribing exercise
. Med Sci Sports Exerc. 2011; 43 (7): 1334–59.
11. Hamer M. The relative influences of fitness and fatness on inflammatory factors. Prev Med. 2007; 4: 3–11.
12. Haskell WL, Lee I-M, Pate RR, et al.. Physical activity and public health: updated recommendation for adults from the American College of Sports Medicine and the American Heart Association. Circulation. 2007; 116: 1081–93.
13. Kasapis C, Thompson PD. The effects of physical activity on serum C-reactive protein
and inflammatory markers: a systematic review. J Am Coll Cardiol. 2005; 45: 1563–9.
14. Kelley GA, Kelley KS. Effects of aerobic exercise
on C-reactive protein
, body composition, and maximum oxygen consumption in adults: a meta-analysis of randomized controlled trials. Metabolism. 2006; 55: 1500–7.
15. 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): 575–81.
16. Kodama S, Saito K, Tanaka S, et al.. Cardiorespiratory fitness as a quantitative predictor of all-cause mortality and cardiovascular events in healthy men and women: a meta-analysis. JAMA. 2009; 301: 2024–35.
17. LaMonte MJ, Durstine JL, Yanowitz FG, et al.. Cardiorespiratory fitness and C-reactive protein
among a tri-ethnic sample of women. Circulation. 2002; 106: 403–6.
18. Majka DS, Chang RW, Vu TH, et al.. Physical activity and high-sensitivity C-reactive protein
: the multi-ethnic study of atherosclerosis. Am J Prev Med. 2009; 36: 56–62.
19. Manson JE, Greenland P, LaCroix AZ, et al.. Walking compared with vigorous exercise
for the prevention of cardiovascular events in women. N Engl J Med. 2002; 347: 716–25.
20. Manson JE, Hu FB, Rich-Edwards JW, et al.. A prospective study of walking as compared with vigorous exercise
in the prevention of coronary heart disease in women. N Engl J Med. 1999; 341: 650–8.
21. Marcell TJ, McAuley KA, Traustadottir T, Reaven PD. Exercise
training is not associated with improved levels of C-reactive protein
or adiponectin. Metabolism. 2005; 54: 533–41.
22. McTiernan A, Kooperberg C, White E, et al.. Recreational physical activity and the risk of breast cancer in postmenopausal women: the Women’s Health Initiative Cohort Study. JAMA. 2003; 290: 1331–6.
23. Meyer AM, Evenson KR, Morimoto L, Siscovick D, White E. Test–retest reliability of the Women’s Health Initiative physical activity questionnaire. Med Sci Sports Exerc. 2009; 41 (3): 530–8.
24. Mora S, Cook N, Buring JE, Ridker PM, Lee I-M. Physical activity and reduced risk of cardiovascular events: potential mediating mechanisms. Circulation. 2007; 116: 2110–8.
25. Poole C. Multiple comparisons? No problem! Epidemiology
. 1991; 2: 241–3.
26. Rossouw J E, Anderson GL, Prentice RL, et al.. Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results From the Women’s Health Initiative randomized controlled trial. JAMA. 2002; 288: 321–33.
27. Selvin E, Paynter NP, Erlinger TP. The effect of weight loss on C-reactive protein
: a systematic review. Arch Intern Med. 2007; 167: 31–9.
28. Smith JK, Dykes R, Douglas JE, Krishnaswamy G, Berk S. Long-term exercise
and atherogenic activity of blood mononuclear cells in persons at risk of developing ischemic heart disease. JAMA. 1999; 281: 1722–7.
29. Stevens J, Truesdale KP, McClain JE, Cai J. The definition of weight maintenance. Int J Obes (Lond). 2006; 30: 391–9.
30. Stewart LK, Earnest CP, Blair SN, Church TS. Effects of different doses of physical activity on C-reactive protein
among women. Med Sci Sports Exerc. 2010; 42 (4): 701–7.
31. Tanasescu M, Leitzmann MF, Rimm EB, Willett WC, Stampfer MJ, Hu FB. Exercise
type and intensity in relation to coronary heart disease in men. JAMA. 2002; 288: 1994–2000.
32. The Women’s Health Initiative Study Group. Design of the Women’s Health Initiative clinical trial and observational study. Control Clin Trials. 1998; 19: 61–109.
33. Tisi PV, Hulse M, Chulakadabba A, Gosling P, Shearman CP. Exercise
training for intermittent claudication: does it adversely affect biochemical markers of the exercise
-induced inflammatory response? Eur J Vasc Endovasc Surg. 1997; 14: 344–50.
34. US Department of Health & Human Services. 2008 Physical Activity Guidelines Advisory Committee report [cited 2011 July 27]. Available from http://www.health.gov/paguidelines/
36. Warnberg J, Cunningham K, Romeo J, Marcos A. Physical activity, exercise
and low-grade systemic inflammation. Proc Nutr Soc. 2010; 69: 400–6.