C-reactive protein (CRP), serum amyloid A (SAA), and interleukin 6 (IL-6) are nonspecific markers of chronic low-grade inflammation that correlate with adiposity, sedentary lifestyle, and low aerobic fitness (5,9,16,20,23). Observational studies have suggested that a reduction in chronic inflammation, often achieved through lifestyle modification, is associated with reduced risk of disease and mortality (1,22,25).
Randomized trials of combined exercise and dietary interventions have been shown to reduce CRP and IL-6 (3,7,14,26), although we are unaware of comparable studies that have examined the combined effects of exercise and dietary intervention on SAA. The effect of exercise alone on inflammatory markers is not clear (2,8,10,12,13,17,21), with a recent meta-analysis suggesting a nonsignificant 3% decrease in CRP from five studies with aerobic exercise interventions of 8 wk or longer (13). It is important to determine the independent effect of exercise on chronic inflammation to better define the public health role of physical activity, in the absence of weight loss or diet intervention, for primary disease prevention.
With inconsistent evidence from observational studies and intervention trials, the role of exercise on reducing chronic inflammation remains unclear. Similarly, the potential for a mediating effect by weight loss is not well understood. In a randomized controlled trial of 115 previously sedentary postmenopausal, overweight or obese women, with high retention, we examined the effect of a 1-yr aerobic exercise intervention compared with a stretching control on CRP, SAA, and IL-6. We hypothesized that the exercise intervention would result in decreased levels of inflammation, and this effect would be partly mediated by fat loss.
METHODS AND PROCEDURES
Setting and Participants
Participants (n = 115) were a subset of women (n = 173) recruited for an exercise intervention trial who met additional eligibility criteria for this study of immune function (4) and inflammation. For this study, 53 exercisers and 62 controls were eligible from the parent trial (Fig. 1). Eligibility criteria included the following: age 50-75 yr; body mass index (BMI) between 25 and 40 kg·m−2 (or 24.0-24.9 kg·m−2 if body fat >33%); postmenopausal; not taking postmenopausal hormones; nonsmoker; sedentary at baseline (<60 min·wk−1 of moderate- and vigorous-intensity recreational activities and V˙O2max <25.0 mL·kg−1·min−1); alcohol consumption of fewer than two drinks per day; no personal history of invasive cancer, diabetes, cardiovascular disease, or asthma; no current serious allergies; no regular (two times per week or more) use of aspirin or other nonsteroidal anti-inflammatory medications; and no use of corticosteroids or other medications known to affect immune function. Women were recruited through a combination of mass mailings and media placements, as described previously (24). Randomization was stratified by BMI (<27.5 or ≥27.5 kg·m−2) to ensure equal numbers of heavier and lighter women in each study group. Approximately equal numbers of exercisers and controls were enrolled during each month of recruitment. All women provided written informed consent, and all procedures were approved by the Fred Hutchinson Cancer Research Center Institutional Review Board.
Randomization and Exercise Intervention
The exercise intervention was designed to progress to at least 45 min of moderate-intensity exercise at a target heart rate (HR) of 60%-75% observed maximal HR, 5 d·wk−1, by the eighth week of the trial. Participants attended three supervised sessions per week at a study facility and exercised 2 d·wk−1 at home during months 1-3. During months 4-12, participants attended at least one session per week at a study facility and exercised 4 d·wk−1 either at home or at a facility. Participants wore Polar HR monitors during all exercise sessions.
V˙O2max was directly assessed at baseline (before randomization) and at 12 months in all participants by maximal-graded treadmill test. Exercise intervention participants kept daily activity logs of all sport and recreational activities of moderate-to-vigorous intensity (estimated to be ≥3 METs). For each exercise session, participants recorded the mode and duration of exercise, peak HR, and relative perceived exertion. Activity logs were reviewed weekly by study staff to monitor compliance and to intervene when needed. Women randomized to the control group attended once-weekly 45-min stretching sessions and were asked to not otherwise change exercise habits during the study. All participants were asked to maintain their usual diet.
At baseline (before randomization), at 3 months, and at 12 months, all women came to the University of Washington (UW) Medical Center for blood draws between 7:30 and 8:30 a.m., after a 12-h fast, under the following blood-draw criteria: no infection or symptoms of any infection for ≥7 d; adequate sleep (6-9 h); no exercise or alcohol for 24 h; no topical corticosteroids or aspirin for 48 h; no systemic antihistamines or corticosteroids for 1 wk; and no immunizations during the previous 3 wk. All participants were contacted 1 wk after their blood draw to track recent illness; several reported being ill shortly after their blood draw and returned for a second blood draw when symptom-free. Serum and plasma were collected and stored at −70°C. All inflammation marker assays were conducted at the UW Clinical Immunology Laboratory (M.H.W.); participant samples from each of the three time points were assayed in the same batch to avoid biased measurement from interassay variability.
C-reactive protein and serum amyloid A.
CRP and SAA were measured in serum by latex-enhanced nephelometry using high-sensitivity assays on the Behring Nephelometer II analyzer (Dade-Behring Diagnostics, Deerfield, IL) with lower detection limits of 0.2 mg·L−1 for CRP and 0.7 mg·L−1 for SAA. Interassay coefficients of variation were 5%-9% for CRP and 4%-8% for SAA.
IL-6 was assayed in duplicate in serum by an ultrasensitive solid-phase sandwich enzyme-linked immunosorbent assay (ELISA) with the Biosource Human IL-6 Immunoassay kit (Biosource, Camarillo, CA).
Body composition and distribution.
At baseline, at 3 months, and at 12 months, body weight to the nearest 0.1 kg and height to the nearest 0.1 cm were taken in duplicate, and the average was used to compute BMI (kg·m−2). At baseline and at 12 months, waist circumference (WC) was measured to the nearest 0.1 cm in duplicate and averaged. Also at baseline and at 12 months, total body fat and percentage body fat were analyzed by DXA (QDR 1500; Hologic, Waltham, MA); and intra-abdominal fat images were obtained via a one-slice computed tomography at L4-L5 (CT; Model CT 9800 scanner; General Electric, Waukesha, WI), with coefficients of variation of 1.2% for both.
Other study measures.
At baseline, at 3 months, and at 12 months, data on demographic information, medical history, health habits, medication use, reproductive and body weight history, total energy intake (via 120-item self-administered food frequency questionnaire), and frequency, duration, and intensity of physical activity (via self-administered Minnesota Physical Activity Questionnaire) were collected.
Intervention effects were assessed on the inflammation- related outcomes at 3 and 12 months after randomization, on the basis of comparisons between exercisers and controls as defined at randomization (i.e., intent-to-treat). Changes between groups were assessed with generalized estimating equations (GEE), which account for repeated observations on the same subjects over time. Baseline measures were included in each GEE model. In secondary analyses, we examined subgroup effects according to baseline values for BMI (above or below 30 kg·m−2) and WC (above or below 88 cm). Additional subgroup analyses were conducted for changes at 12 months in aerobic fitness (controls vs exercisers who increased V˙O2max by <5%, 5%-15%, or >15%), exercise adherence (controls vs exercisers who performed physical activity an average of <136, 136-195, or >195 min·wk−1), and 12-month changes in percentage body fat (controls vs exercisers who had minimal percent fat loss (<0.1%) or gained percent body fat, had lost between 0.1% and ≤2% body fat, or had lost >2% body fat), body weight (controls vs exercisers who had minimal weight loss (<0.5 kg) or gained body weight, had lost between 0.5 and ≤3 kg body weight, or had lost >3 kg of body weight), WC (controls vs exercisers who had minimal reduction in WC (<0.5 cm) or gained WC, had lost between 0.5 and ≤3 cm of WC, or had reduced WC by >3 cm), and intra-abdominal fat (controls vs exercisers who had gained intra-abdominal fat or had minimal reduction in intra-abdominal fat (<2 cm2), had decreased intra-abdominal fat by 2-8 cm2, or had decreased intra-abdominal fat by >8 cm2). CRP, SAA, and IL-6 were non-normally distributed and were log-transformed for all analyses. Missing data were omitted from GEE analyses. All statistical tests were two-sided. All analyses were performed with SAS software (version 9.1; SAS Institute, Cary, NC).
A total of 115 women were randomized: 53 to intervention and 62 to control. One exerciser did not have data for CRP and SAA at baseline and was excluded. Two participants (one exerciser, one control) were excluded from analyses because of unusually high CRP and SAA concentrations at one time point in the study (i.e., approximately 10-fold greater than their values at other time points in the study, suggesting an unreported or unknown acute inflammatory event). Of the 112 remaining participants at baseline (exercisers, n = 51; controls, n = 61), 106 (95%) provided blood samples at 3 months (exercisers, n = 48; controls, n = 58), and 104 (93%) provided blood samples at 12 months (exercisers, n = 47; controls, n = 57) (Fig. 1).
At baseline, there were no significant differences between exercisers and controls (Table 1). Women were, on average, 60 yr of age, and most described themselves as non-Hispanic white. In line with the eligibility criteria, participants had an average BMI of 30 kg·m−2, 47% body fat, and low aerobic fitness. For 12 months, the intervention group participated in moderate activity an average of 3.8 d·wk−1 (SD 1.3) for 166 min·wk−1 (SD 76.1). At the end of study, exercisers increased aerobic fitness (V˙O2max) by 13.8%, whereas controls experienced a 0.1% increase (between-group P value at 12 months <0.0001). Similarly, the exercise intervention resulted in decreased body weight versus control (exercise: −1.8 kg; control: +0.3 kg; P = 0.002) and decreased percentage body fat (exercise: −1.5%; control: +0.02%, P < 0.0001). There were no differences between or within intervention groups for total caloric intake throughout the trial (baseline values are shown in Table 1; 12-month values: exercisers 1618 ± 621 kcal; controls 1684 ± 720 kcal, all P ≥ 0.47).
Main intervention effects.
Exercisers statistically significantly decreased CRP relative to controls at 12 months (Table 2). SAA decreased among exercisers and increased slightly among controls, although the between-group difference was not statistically significant. IL-6 was not affected by the exercise intervention. No intervention effects were noted at 3 months. Results were similar when statin users (n = 7) were excluded.
Stratified results: baseline BMI and WC.
The effect of exercise intervention on CRP was restricted to women who were obese at baseline (Table 3). Obese women (BMI ≥ 30 kg·m−2) in the exercise group experienced a statistically significant and continuous decline in CRP during the 12-month trial (baseline CRP: 3.95 mg·L−1 (95% CI, 2.87-5.44); 3-month CRP: 3.65 mg·L−1 (95% CI, 2.6-5.14); 12-month CRP: 3.16 mg·L−1 (95% CI, 2.38-4.2)); whereas obese women in the control group experienced a moderate increase in CRP values during the same period (P value for between-group difference at 12 months = 0.002). Similar trends were noted for women with abdominal obesity (P < 0.0001). BMI and WC at baseline were highly correlated (r = 0.82; P < 0.0001), and both anthropometric measures were identically correlated with baseline CRP (both correlations = 0.42, P < 0.0001), which explains these similar trends. Similar trends across BMI and WC were observed for SAA, although the effects only reached borderline statistical significance for the obese BMI group (P = 0.08).
Intervention effects on CRP stratified by exercise adherence and fat loss.
The exercise intervention group was stratified into subgroups on the basis of 12-month changes in aerobic fitness (percentage change in V˙O2max), average adherence to the exercise intervention (physical activity minutes per week), and 12-month changes in percentage body fat, body weight, WC, and intra-abdominal fat. As described above, among women in the control group, CRP concentrations increased during the 12-month trial. Linear trends were observed between CRP and 12-month changes in V˙O2max (P = 0.006), exercise adherence (P = 0.004), body fat loss (P = 0.001), weight loss (P = 0.002), WC loss (P = 0.02), and intra-abdominal fat loss (P = 0.03) (Table 4). No subgroup effects were noted for SAA or IL-6 (data not shown).
This study demonstrated that a yearlong aerobic exercise intervention compared with a stretching control program reduced CRP among previously sedentary obese postmenopausal women. The obese women in this study had clinically "high" average CRP values according to established criteria (19), and despite decreased CRP after 3 and 12 months of exercise intervention, their CRP values remained high throughout the trial. Future studies should assess exercise interventions of longer duration and greater intensity to determine the dose of exercise or fat loss needed to reduce CRP further. Although the CRP values in this study remained relatively high despite an exercise intervention effect, obese exercisers experienced a decrease in CRP of an average of 1.59 mg·L−1 for 12 months, which may have public health relevance in that a recent prospective study reported that each 1.02-mg·L−1 increase in CRP was associated with approximately a 35% increase (95% CI, 1.05-1.74) in colon cancer risk (6). Similarly, a recent cohort study identified a 29% increase (95% CI, 1.07-1.55) in coronary heart disease risk for a 1-unit increase in log CRP (11).
Our stratified analyses showed that only those exercisers who decreased body fat by 2% or greater experienced a statistically significant reduction in CRP versus controls, suggesting that the exercise effect is at least partially dependent on fat loss. Moreover, the exercisers who lost more body weight/body fat were largely the same exercisers who were most adherent to the exercise intervention (χ2 = 10.34, P = 0.03). And because weight/fat loss was generally minimal in this trial, this suggests that exercise, in the absence of caloric restriction, is sufficient in reducing CRP among obese women. Two previous multiarm randomized controlled trials have offered valuable perspective concerning the role of exercise and diet interventions (both alone and in combination) on biomarkers of inflammation (18,27). Among a group of 316 community-dwelling older adults with movement limitations, diet alone (but not exercise alone or diet and exercise) favorably reduced CRP and IL-6 (18). Among a study population of 34 sedentary, overweight or obese, postmenopausal women (i.e., a population that is more comparable to the current study), a 6-month exercise-and-diet intervention, but not diet alone, decreased CRP and IL-6 (27). Given the somewhat discrepant findings and small number of studies on this topic, more multiarm trials seem justified to better discern the individual and combined effects of diet and exercise on inflammation.
A recent meta-analysis of five randomized controlled trials summarized the effect of aerobic exercise alone on CRP (13). The studies included in the meta-analysis had intervention durations of 8 wk to 6 yr and were undertaken in varied populations, including individuals with rheumatoid arthritis (2), postmenopausal breast cancer survivors (8), elderly men and women (10), overweight men and women (17), and middle-aged men (21). Overall, a statistically nonsignificant 3% decrease in CRP was reported; however, most individual studies suggested a trend toward exercise-driven reduction of CRP (8,10,21). All but one (21) of these previous trials had considerably shorter intervention durations than the current study, and given that our 3-month analyses suggested no exercise effect, the collective evidence suggests that longer periods of exercise are required to reduce CRP.
The 6-yr trial by Rauramaa et al. (21) showed a modest trend toward reduction in CRP; compared with the women in the current study, however, those male participants were considerably leaner, more aerobically fit, and had lower baseline CRP values. Consistent with the present study, a more recent trial (not included in the above meta-analysis) suggested that 6 months of exercise alone statistically significantly reduced CRP among obese patients with type 2 diabetes mellitus despite minimal changes in body weight (12). Similarly, a recent randomized trial of 10 months of aerobic exercise versus flexibility/strength training reduced CRP, IL-6, and IL-18 among older men and women (15). From these data, it seems that exercise is most beneficial in reducing chronic inflammation among those persons with relatively high levels of baseline inflammation.
The specific pathways through which exercise may decrease CRP are not established, but several plausible mechanisms exist. One mechanism is through regulation of interleukins and related cytokines, including IL-6 and tumor necrosis factor α, both of which are released by adipose tissue (as well as by peripheral blood mononuclear cells), and the former stimulates hepatic release of CRP. Because our study showed no effect of exercise on serum IL-6, our data do not support the direct role of circulating IL-6 on CRP reduction.
Strengths of the current study include the relatively long duration of exercise intervention, good adherence and high retention rates, effective aerobic exercise intervention as demonstrated by gold standard measures of cardiorespiratory fitness, and a randomized controlled trial design. Limitations that should be considered when interpreting these data include the strict inclusion criteria that preclude generalization to other populations and the narrow range of inflammatory markers selected. Secondly, the intervention was not intended to induce weight loss, and we cannot therefore assess the effect of substantial weight loss, with or without exercise, on chronic inflammation. Lack of racial diversity is another limitation of the current study that should be addressed in future intervention trials; in a cross-sectional data, LaMonte et al. (16) showed that cardiorespiratory fitness was correlated with CRP among white and Native American, but not African American, women.
In conclusion, a 12-month aerobic exercise intervention resulted in improved aerobic fitness and reduced CRP, despite relatively little fat loss. CRP reduction was limited to women who were obese at baseline, a group with clinically high CRP values.
This study was supported by research grants from the National Institutes of Health (NIH; CA 69334, DK 02860, and DK 035816). Dr. P. Campbell was a Research Fellow of the Canadian Cancer Society through an award from the National Cancer Institute (NCI) of Canada and also supported by a NIH Transdisciplinary Research on Energetics and Cancer Postdoctoral Fellowship (NCI U54 CA116847). Dr. K. Campbell was supported through a Research Fellow Award from the Canadian Institutes for Health Research. The sponsors had no role on the design or conduct of the study, the collection, the management, the analyses, the interpretation of the data, or on preparing, reviewing, or approving the manuscript.
Disclosure statement: The authors have no conflicts to disclose.
The results of this study do not constitute endorsement by the ACSM.
Current affiliation for Dr. P. Campbell: American Cancer Society, Atlanta, GA.
1. Alley DE, Crimmins E, Bandeen-Roche K, Guralnik J, Ferrucci L. Three-year change in inflammatory markers in elderly people and mortality: the Invecchiare in Chianti study. J Am Geriatr Soc
2. 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
3. Bo S, Ciccone G, Baldi C, et al. Effectiveness of a lifestyle intervention on metabolic syndrome. A randomized controlled trial
. J Gen Intern Med
4. Campbell PT, Wener MH, Sorensen B, et al. Effect of exercise on in vitro
immune function: a 12-month randomized, controlled trial among postmenopausal women. J Appl Physiol
5. 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
6. Erlinger TP, Platz EA, Rifai N, Helzlsouer KJ. C-reactive protein
and the risk of incident colorectal cancer. JAMA
7. Esposito K, Pontillo A, Di Palo C, et al. Effect of weight loss and lifestyle changes on vascular inflammatory markers in obese women: a randomized trial. JAMA
8. Fairey AS, Courneya KS, Field CJ, Mackey JR. Physical exercise and immune system function in cancer survivors: a comprehensive review and future directions. Cancer
9. 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
10. Hammett CJ, Oxenham HC, Baldi JC, et al. Effect of six months' exercise training on C-reactive protein
levels in healthy elderly subjects. J Am Coll Cardiol
11. Hung J, Knuiman MW, Divitini ML, Langton PE, Chapman CL, Beilby JP. C-reactive protein
and interleukin-18 levels in relation to coronary heart disease: prospective cohort study from Busselton Western Australia. Heart Lung Circ
12. Kadoglou NP, Iliadis F, Angelopoulou N, et al. The anti-inflammatory effects of exercise training in patients with type 2 diabetes mellitus. Eur J Cardiovasc Prev Rehabil
13. 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
14. Kim YJ, Shin YO, Bae JS, et al. Beneficial effects of cardiac rehabilitation and exercise after percutaneous coronary intervention on hsCRP and inflammatory cytokines in CAD patients. Pflugers Arch
15. Kohut ML, McCann DA, Russell DW, et al. Aerobic exercise, but not flexibility/resistance exercise, reduces serum IL-18, CRP, and IL-6 independent of beta-blockers, BMI, and psychosocial factors in older adults. Brain Behav Immun
16. LaMonte MJ, Durstine JL, Yanowitz FG, et al. Cardiorespiratory fitness and C-reactive protein
among a tri-ethnic sample of women. Circulation
17. Marcell TJ, McAuley KA, Traustadottir T, Reaven PD. Exercise training is not associated with improved levels of C-reactive protein
or adiponectin. Metabolism
18. Nicklas BJ, Ambrosius W, Messier SP, et al. Diet-induced weight loss, exercise, and chronic inflammation
in older, obese adults: a randomized controlled clinical trial. Am J Clin Nutr
19. 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
20. Pischon T, Hankinson SE, Hotamisligil GS, Rifai N, Rimm EB. Leisure-time physical activity
and reduced plasma levels of obesity-related inflammatory markers. Obes Res
21. Rauramaa R, Halonen P, Vaisanen SB, et al. Effects of aerobic physical exercise on inflammation
and atherosclerosis in men: the DNASCO Study: a six-year randomized, controlled trial. Ann Intern Med
22. Ridker PM, Hennekens CH, Buring JE, Rifai N. C-reactive protein
and other markers of inflammation
in the prediction of cardiovascular disease in women. N Engl J Med
23. Rothenbacher D, Hoffmeister A, Brenner H, Koenig W. Physical activity
, coronary heart disease, and inflammatory response. Arch Intern Med
24. Tworoger SS, Yasui Y, Ulrich CM, et al. Mailing strategies and recruitment into an intervention trial of the exercise effect on breast cancer biomarkers. Cancer Epidemiol Biomarkers Prev
25. Tzoulaki I, Murray GD, Lee AJ, Rumley A, Lowe GD, Fowkes FG. C-reactive protein
, interleukin-6, and soluble adhesion molecules as predictors of progressive peripheral atherosclerosis in the general population: Edinburgh Artery Study. Circulation
26. Villareal DT, Miller BV 3rd, Banks M, Fontana L, Sinacore DR, Klein S. Effect of lifestyle intervention on metabolic coronary heart disease risk factors in obese older adults. Am J Clin Nutr
27. You T, Berman DM, Ryan AS, Nicklas BJ. Effects of hypocaloric diet and exercise training on inflammation
and adipocyte lipolysis in obese postmenopausal women. J Clin Endocrinol Metab