Exercise-induced changes in inflammatory processes: Implications for thrombogenesis in cardiovascular disease

Abstract

Sedentary lifestyle is a risk factor and a strong predictor for chronic disease and premature death. Low-grade inflammation has been proved a key player in the pathogenesis of cardiovascular disease. Inflammatory processes have been also involved in maintaining the balance between coagulation and fibrinolysis. In addition, an inverse linear dose–response relation between physical activity and mortality risks has also been reported. However, the favorable effects of structured exercise programs and the independent contribution of physical activity to cardiovascular risk are still under investigation.

In response to heavy exercise, interleukin-6 (IL-6) is secreted by contracting skeletal muscles, followed by an acute reactant release of C-reactive protein (CRP). Both CRP and IL-6 can stimulate monocyte tissue factor production, provoke platelet hyperreactivity, promote fibrinogen biosynthesis, and enhance microparticle formation and erythrocyte aggregability, thus triggering prothrombotic state. By contrast, regular exercise and physical activity are protective against all-cause mortality through suppressing pro-inflammatory cytokine production, enhancing anti-inflammatory mediators and antioxidant development, and promoting fibrinolytic activity. Low-load resistance exercise also plays an advantageous role in thrombogenesis by reducing inflammatory processes and potentiating fibrinolytic features.

In the present review article, we provide an overview of the impact of different modes and intensities of physical activity on vascular inflammation and thrombogenesis.

Key words::

• Erythrocyte aggregability
• exercise
• inflammation
• leukocyte activation
• thrombosis

Key messages

• Sedentary lifestyle is a risk factor and a strong predictor for chronic disease and premature death. An inverse linear dose–response relation between physical activity and mortality risks has also been reported.

• In response to heavy exercise, inflammation can stimulate monocyte tissue factor production, provoke platelet hyperreactivity, promote fibrinogen biosynthesis, and enhance microparticle formation and erythrocyte aggregability, thus triggering prothrombotic state.

• By contrast, regular exercise and physical activity are protective against all-cause mortality through suppressing pro-inflammatory cytokine production, enhancing anti-inflammatory mediators and antioxidant development, and promoting fibrinolytic activity. Low-load resistance exercise also plays an advantageous role in thrombogenesis by reducing inflammatory processes and potentiating fibrinolytic features.

Introduction

Cardiovascular disease (CVD) is becoming increasingly common and remains the leading cause of death in modern societies. Independently of body mass index (BMI), sedentary lifestyle is a risk factor (1) and a strong predictor (2) for chronic disease and premature mortality. A specific cluster of ‘diseasome’ attributable to physical inactivity (3), including type 2 diabetes (4) and CVD (5), is closely associated with systemic inflammation.

Given that low-grade inflammation as a key player in the pathogenesis of atherosclerosis (6), how these inflammatory markers affect thrombogenesis (7) and blood rheology (8,9) may provide an impression on the progress or end-stage of the disease. The presence of excess lipoprotein, vascular damage, and increased systemic inflammation can shift the intravascular environment to a hypercoagulable state, subsequently leading to thrombosis and vascular clotting (6,7). The most commonly investigated inflammatory cytokines, such as C-reactive protein (CRP) and interleukin-6 (IL-6), promote a procoagulant environment by up-regulating prothrombotic proteins and down-regulating antithrombotic molecules (7,10). Both IL-6 (11) and CRP (12) can stimulate tissue factor (TF) production from monocyte, consequently triggering the extrinsic coagulation pathway. Moreover, inflammation-induced oxidative stress can also augment leukocyte activation (6), increase platelet production and their sensitivity to thrombin (13), promote fibrinogen biosynthesis (14), and enhance microparticle release (15) and nearby erythrocyte aggregability (16). Therefore a vicious circle linking coagulation, inflammation, and thrombosis appears to be closely associated with atherosclerosis (17,18).

Numerous large, epidemiological studies have linked people who have better cardiovascular fitness or engage in regular physical activity to CVD risk or all-cause mortality reduction (19–24). An inverse linear dose–response relation between the ‘volume of physical activity’ (i.e. intensity, duration, and frequency) and mortality risk has also been well established (25,26). However, self-reported discrepancy of physical activity levels and the generalized regular training protocols cannot provide a timely reflection of the favorable exercise prescriptions on inflammatory processes and subsequent thrombotic changes.

The aim of this review article is to provide an overview of the effects of exercise intensity, duration, and mode on vascular inflammation and discuss their implications in thrombogenesis and, subsequently, in the development of CVD. Moreover, comprehensive advice related to exercise training volume (i.e. intensity × duration; in the unit of metabolic equivalent task (MET)) at a weekly basis will be given as clinical implication references.

Search strategy

A literature review was conducted using computerized databases, including PubMed, MEDLINE, Cochrane Controlled Trials Registry, and EMBASE. The search keywords relating to exercise, inflammation, and thrombogenesis were utilized as follows: The terms ‘exercise’, or ‘physical activity’, or ‘exercise training’, or ‘aerobic exercise’, or ‘resistance exercise’, or ‘endurance exercise’, or ‘interval training’ were combined using AND with the terms ‘inflammation’, or ‘cytokine’, AND with the terms ‘thrombogenesis’, or ‘thrombosis’, or ‘coagulation’, or ‘fibrinolysis’, or ‘platelet’, or ‘erythrocyte’.

Inclusion criteria

We included controlled clinical trials, randomized control trials, comparative studies, acute exposure to an exercise bout, or multicenter investigations measuring the effects of exercise on inflammation and thrombogenesis. The following criteria were limited to human subjects and articles that were published in a peer-reviewed journal.

Exclusion criteria

Animal studies, non-English papers, studies in which the effects of exercise were confounded by other factors such as diet or pharmacological intervention were excluded. Unpublished theses were also excluded. After searching, we considered 127 suitable publications. Additional citations were searched from the reference lists of retrieved papers, review articles, or meta-analyses. Finally, we reviewed 145 references in total.

Exercise and inflammation

A sedentary lifestyle is associated with elevated basal circulating levels of IL-6 and CRP. It has been experimentally demonstrated that skeletal muscle contraction-induced cytosolic Ca²⁺ elevation and mitogen-activated protein kinase (MAPK) production led to transcription factor activation and subsequent IL-6 secretion (3). The favorable effects of physical exercise on several traditional cardiovascular risk factors, such as blood pressure levels, weight control, and lipid profile, have been thoroughly assessed in series of large-scale epidemiological studies regardless of ethnicity (23). However, with respect to cardiovascular disease and low-grade chronic vascular inflammation, the underlying mechanisms of exercise-related benefits are still unclear.

One possible pathway of physical activity benefits may be ascribed to its anti-inflammatory effects. Regular exercise inhibits NF-κB signaling and thereby offers protection against TNF-α-induced insulin resistance. Moreover, the cytokine response to exercise does not include TNF-α and IL-1 but only an increase in IL-6, followed by serial anti-inflammatory cytokines release (e.g. IL-1ra, TNF-R, and IL-10). However, the studies that assessed the anti-inflammatory effects of exercise are characterized by considerable inconsistency (27).

There are many studies on the effects of exercise on the regulation of inflammatory mediators (Table I). Several cross-sectional studies have shown concordant results towards a pro-inflammatory state after acute intense exercise (29,30,32,33,35) and an anti-inflammatory condition after long-term moderate training (36–39,42,46–49,51–53,55,56,60–62). However, the studies lack comprehensive evaluation of both pro- and anti-inflammatory processes and thus provide only fragmentary data on the potential responses of physical activity. Meanwhile, some reports have yielded conflicting data due to diversions in exercise protocols (e.g. mode, intensity, and frequency) (31,34,41–44,46,58,59,62), observation timings (e.g. during the exercise period or in the recovery phase) (30,35,45,53), and studied populations (e.g. age, gender, health conditions) (28,36,37,63). Therefore the methodological discrepancies in evaluating effects of exercise need to be borne in mind.

Table I. Studies of exercise effects on inflammatory processes.

Source	Subjects	Exercise	Main results	Summary
Acute bout exercise
Aerobic
Mendham et al., 2012 (32)	10 sedentary male, randomized cross-over to aerobic exercise and rugby	Cycle ergometry and modified rugby: 40 min, separated by 7 days’ recovery	↑ IL-6 and IL-1ra, ↓ HOMA within 240 min post-exercise period, without differences between protocols	Acute cycle ergometry and rugby have similar inflammatory and glucose regulatory responses
Andersson et al., 2010 (30)	20 fit but not athletic men	Heavy physical activity: 14 days cross-country skiing tour, 12–30 km/day (10 hours)	Exercise ↑ CRP and TNF-α; CRP levels were lower during recovery compared to baseline	Heavy physical exercise ↑ inflammatory variables
Palmer-Kazen et al., 2009 (29)	19 IC patients; 5 IC patients without calf pain (control group)	Treadmill exercise test	Exercise ↑ IL-6, ↓ plasma FGF-2, and ↑ biopsies VEGF-A mRNA	Exercise in IC triggers inflammatory and angiogenesis responses
Harris et al., 2008 (28)	16 overweight men: 8 active (VO2peak = 34.2 ± 1.7) and 8 inactive (VO2peak = 30.9 ± 1.2)	3 different intensity treadmill tests: low (25% VO2peak); moderate (50% VO2peak); high (75% VO2peak), each for 45 min, separately at least 2 days apart	Independent of exercise intensity, FMD ↑ 24% in active group, but ↓ 32% in inactive group following acute exercise. Moderate and high intensities ↑ IL-6 in both groups; TNF-α unchanged	Physical activity is an independent determinant of the FMD response to acute exercise between active and inactive men
Kop et al., 2008 (33)	36 CAD inpatients after successful PCI; 28 healthy controls	Mental challenge tasks (anger recall and mental arithmetic) and treadmill exercise test (standard Bruce protocol)	Both mental challenge and exercise ↑ CRP in CAD patients than in healthy controls. Exercise ↑↑ CRP, IL-6, and sICAM-1 than mental challenge tasks	Exercise ↑ inflammatory markers in patients with CAD than in healthy subjects, and it is related to neurohormonal stress response
Markovitch et al., 2008 (31)	12 sedentary male completed exercise and rest (sitting) trials 10 days apart	Moderate-intensity treadmill walking: 30 min × 50% VO2max	No significant changes in neutrophils, lymphocytes, monocytes, or IL-6, IL-10, CRP over the 7 days following exercise	A single bout of moderate exercise did not change inflammatory status in middle-aged men
Li et al., 1999 (128)	15 healthy young men	Exhaustive exercise test. With or without 1 week aspirin pretreatment	↑ platelet P-selectin expression; ↑ neutrophil and lymphocyte CD11b expression; ↑ platelet–platelet and platelet–leukocyte aggregates; ↑ soluble P-selectin expression, F1 + 2	Acute strenuous exercise induced platelet and leukocyte activation and aggregation. Aspirin treatment attenuated certain platelet activity, but did not change the prothrombotic effects of exercise
Resistance
Vella et al., 2011 (35)	10 recreational active young subjects, randomized to exercise (n = 5) or control (n = 5) group	Intense resistance exercise: 3 sets × 8–12 repetitions, 80% 1RM. Muscle biopsy	2 h after exercise ↓ IκB-α, ↑ phosphorylated NF-κB (p65), ↑ MCP-1, IL-6, IL-8 mRNA and returned to near baseline by 4 h post-exercise	Single bout of intense resistance exercise ↑ NF-κB signaling-mediated inflammatory response in vastus lateralis during the first 2 h after exercise
Barnes et al., 2010 (34)	27 men, randomized to large (n = 11) or small muscle (n = 16) group	Eccentric exercise: bilateral leg press (large muscle, 1RM) or unilateral elbow flexor (small muscle, 3 s per contraction × 2 sets × 20 repetitions)	Eccentric exercise ↓ isometric strength, ↑ muscle soreness, and ↑ cfPWV after 48 hours. Leg exercise ↑CRP; arm exercise ↑ CK	Eccentric exercise transiently ↑ central arterial stiffness, but only the results in small muscle exercise are associated with muscle damage
Regular exercise or training
Aerobic
Czepluch et al., 2011 (62)	13 young, healthy individuals	3 weeks’ strenuous exercise training: 2 sessions × 6 days per week, 50 min ergometer cycling, and 40 min running indoors, at lactate threshold intensity (64.8–75.6 MET-hours/week)	Exercise training ↓ VEGF-A, TGF-β, MCP-1 and lasts for 4 weeks	Strenuous exercise training ↓ monocyte chemotaxis
Giallauria et al., 2011 (60)	75 post-infarction patients in a single center, randomized to exercise training (n = 37) and control (n = 38) group	6 months’ exercise training: 3 times/week, 30 min ergometer, 60–70% VO2peak (4.2–4.9 MET-hours/week)	Exercise training ↓ HMGB1, LVEDV, and wall motion score index, ↑ VO2peak	Exercise training ↓ HMGB1, ↑ cardiopulmonary and autonomic function, and favorable to cardiac remodeling
Lara Fernandes et al., 2011 (46)	34 CAD patients, randomized to exercise (n = 15) or control (n = 19) group	4 months’ exercise training: 40 min cycling at 65% VO2peak, 3 days/week; an acute exercise test before and after 4 months’ training (9.2 MET-hours/week)	Acute exercise test ↑ CRP, Mig, and VCAM-1. 4 months’ exercise training ↓ BNP at rest and CRP provoked by acute test	Chronic exercise training might partially reverse the inflammatory response caused by acute exercise in CAD patients
Myers et al., 2010 (38)	57 early AAA patients, randomized to exercise training (n = 26) or usual care (n = 31) group	1 year in-house and home moderate exercise training: 60%–80% HRR intensity (RPE 12–14), 45 min treadmill, cycling, stair climbing, elliptical training, and rowing, plus 10 min resistance exercise, 3 times/week	Exercise training ↑ exercise capacity, ↓ CRP, and tend to ↓ waist circumference and waist-to-hip ratio. No ↑ AAA growth rate or adverse clinical events while exercise training	Exercise training is well tolerated and sustainable in small AAA subjects over 1 year
Wilund et al., 2010 (47)	17 hemodialysis patients, randomized to exercise (n = 8) or control (n = 9) group	4 months’ cycling training: RPE at 12–14, 45 min/session, 3 days/week	Exercise ↑ 17% shuttle walk performance, ↓ 38% TBARS, ↓ 27% ALP, and ↓ 11% epicardial fat layer	Endurance exercise training may improve CVD risk in hemodialysis patients by ↓ serum oxidative stress (TBARS), vascular calcification risk factor (ALP), and epicardial fat
Arsenault et al., 2009 (41)	349 moderately ↑ SBP, postmenopausal overweight/obese women (45–75 years, BMI = 25–43 kg/m^2), randomized to control (n = 82) or intervention (n = 267, 3 energy expenditures): 4 (n = 117), 8 (n = 70), or 12 (n = 80) kcal/kg/week (KKW) groups	6-month exercise intervention (semi-recumbent cycle ergometer and treadmill): 50% VO2max, 3–4 times/week, training time varied from 72.2 ± 12.3 (4KKW), 135.8 ± 19.5 (8KKW) to 191.7 ± 33.7 (12KKW) min/week (7.9 MET-hours/week)	Exercise ↑ VO2max (training dose-dependent manner), weight loss, and ↓ waist circumference, but no changes in SBP, lipoprotein-lipid profile, CRP, IL-6, TNF-α, and adiponectin	Exercise improved fitness (in dose-dependent manner), not cardiometabolic risk profile in sedentary, but metabolically healthy obese/overweight women
Balen et al., 2008 (48)	60 AMI (7.08 ± 1.60 days) patients (60 ± 10 years), randomized to exercise or control group	3-week moderate-intensity exercise	AMI ↑ inflammatory state. Exercise training ↓ acute phase reactants and TNF-αSR1	Exercise training ↓ AMI-induced inflammatory state
Drexel et al., 2008 (42)	45 healthy sedentary individuals (16 ♂, 29 ♀, mean 48 years), randomized to eccentric or concentric exercise, and crossing over afterward	4 months’ exercise training: altitude 540 m, 3–5 times/week; 2 months of hiking upwards (concentric), and 2 months of hiking downwards (eccentric), with a cross-over	Eccentric exercise ↓ total cholesterol, LDL, apolipoprotein B/A1 ratio, HOMA-IR score, and CRP, ↑ glucose tolerance. Concentric exercise ↓ triglyceride	Eccentric endurance exercise is more favorable for metabolic and anti-inflammatory effects than concentric exercise
Pierce et al., 2008 (56)	14 heart transplantation recipients 8 weeks afterward, randomized to exercise (n = 8) or control (n = 6) group	12-week supervised aerobic exercise training	Exercise training ↑ peak CBF (during reactive hyperemia after 5 min of limb ischemia), ↑ exercise test duration and VO2max. After 12 weeks, only control group ↑ CRP, IL-6, TNF-α, and sICAM-1	Aerobic exercise training ↑ endothelium-dependent vasodilation of the calf, but not forearm resistance arteries, and ↓ pro-inflammatory markers
Pullen et al., 2008 (55)	19 HF patients (NYHA class I–III, mean EF = 25%), randomized to yoga (YT; n = 9) and standard (MT; n = 10) treatment	8 weeks yoga treatment: twice a week, 70 min (6.44 MET-hours/week)	Yoga ↑ GXT time, VO2peak, EC-SOD, and QoL, ↓ IL-6 and hsCRP	Yoga ↑ exercise tolerance, QoL, and ↓ inflammatory markers in HF patients
Walther et al., 2008 (61)	101 male stable CAD (indicated for revascularization), randomized to exercise training (ET; n = 51) or PCI with stent implantation (n = 50)	2 years’ regular exercise training	Both ET and PCI ↑ VO2max, but ET ↓↓ CRP, IL-6, ↑↑ event-free survival rate than PCI	Exercise training ↓ inflammatory markers and ischemic events in stable CAD patients
Peschel et al., 2007 (59)	39 CAD patients with statin treatment, randomized to exercise or conventional therapy group	4 weeks’ submaximal cycle ergometer training (6 × 15 min) plus 1 hour/week group exercise session, followed by 5 months’ 30 min/day home-based ergometer training plus 1 hour/week group exercise	Exercise training ↑ VO2max, ↓ lipids, BMI, BP, fasting glucose, HbA1c, Mac-1, and VLA-4. After 5 months of a home-based intervention, the beneficial effects were blunted	CAD patients might further benefit by higher amount of exercise training in anti-atherogenic effects
Hammett et al., 2006 (45)	152 healthy female smokers, randomized to exercise training or health education group	12 weeks’ exercise training (60%–70% HRmax, 45 min × 3 sessions per week) (9 MET-hours/week)	Exercise ↑ fitness at 6 and 12 weeks. No differences of any inflammatory marker between the exercise and control groups at either 6 or 12 weeks	Regular exercise ↑ fitness, but did not ↓ any of the inflammatory markers
Meyer et al., 2006 (57)	67 obese children, randomized to exercise or control group	6 months’ aerobic exercise training: 60–90 min/day, 3 times/week	Exercise training ↑ FMD and ↓ IMT and cardiovascular risk factors (BMI, body fat mass, waist/hip ratio, ambulatory SBP, fasting insulin, triglycerides, LDL/HDL ratio, CRP and fibrinogen)	Exercise training restores endothelial function and improves carotid IMT in obese children
Gielen et al., 2005 (49)	20 male CHF (EF = 25 ± 2%, age = 54 ± 2 years), randomized to exercise (n = 10) or control (n = 10) group	6 months’ exercise training, 70% VO2max, 10 min/day, 6 days/week (4 MET-hours/week)	Exercise training ↑ 27% cytochrome c oxidase (COX) activity. Changes in iNOS expression and protein content were inversely correlated with changes in COX activity	Exercise training ↑ muscular oxidative metabolism by ↑anti-inflammatory effects
Niebauer et al., 2005 (50)	18 CHF patients; 9 age-matched controls, randomized to exercise or rest group	8 weeks’ exercise training (70%–80% HRmax, 30 min × 5 days per week bicycle ergometer and calisthenics 9 min/day) (11.9 MET-hours/week)	Exercise training ↑ VO2peak in CHF, no activation of TNF-α, TNF-αSR1 and 2, IL-6, sE-selectin, sICAM, sCD14 (neither in patients nor controls)	Chronic exercise does not ↑ pro-inflammatory cytokines and markers of endothelial damage
LeMaitre et al., 2004 (51)	46 stable CHF, randomized to exercise training (n = 24) or electrical stimulation (quadriceps and gastrocnemius muscles) (n = 22) group	6 weeks’ bicycle ergometer exercise training, 30 min × 5 days per week, 70% HRmax	Exercise performance ↑ in both groups. Only bicycle exercise training ↓ TNF-αSR2, and has a decrease tendency of TNF-αSR1, and TNF-α	Active physical training ↑ exercise capacity and ↓ pro-inflammatory response for CHF
McDermott et al., 2004 (40)	25 PAD patients without symptoms of claudication, randomized to exercise (n = 17) or usual care (n = 8) group	12 week treadmill walking program: 3 times/week, 30–50 min/session, begun at 3.2 km/h, 0% incline for 20 min, increased by 0.8 km/h, and 2% incline respectively (RPE 11–12)	Exercise training ↑ WIQ walking speed score, 6-min walk distance, and maximal treadmill walking distance; but no significant changes in the inflammatory blood factors	Supervised treadmill walking program ↑ walking performance without evoking inflammatory markers
Milani et al., 2004 (39)	277 CAD patients: 235 patients, phase II cardiac rehabilitation program; 42 controlled patients, regular medical care	3 months’ group exercise training, 3 times/week, at anaerobic threshold intensity; and 1–3 times/week home-based exercise (7 MET-hours/week)	Rehabilitation patients ↓ body fat, obesity indices, ↓ hsCRP (regardless of statin treatment or weight loss) and other cardiac risk factors, ↑ exercise capacity	Exercise rehabilitation program ↓ cardiac risk factors and CRP (additional benefit beyond statin therapy)
Rauramaa et al., 2004 (44)	100 middle-aged white men, randomized to exercise (n = 51) or control (n = 49) group	6 years’ low-to-moderate intensity aerobic exercise: 40%–60% VO2max, 45–60 min × 5 times per week (16 MET-hours/week)	Exercise group ↑ ventilatory aerobic threshold, ↓ IMT (after adjusting statin, smoking, LDL, SBP, and waist circumference), but no influence on hsCRP	Aerobic exercise training attenuates progression of atherosclerosis in a subgroup of men not taking statins
Gielen et al., 2003 (52)	20 male stable CHF (LVEF = 25% ± 2%; 54 ± 2 years), randomized to exercise (n = 10) or control (n = 10) group	6 months’ exercise training: 70% VO2peak bicycle ergometer, 10 min × 2 weeks; after discharge, 70% VO2peak, 20 min daily cycling plus a 60 min weekly group training (walking, calisthenics, and non-competitive ball games)	Exercise training ↑ VO2peak, ↓ vastus lateralis muscle TNF-α, IL-6, IL-1β, and local iNOS, but serum levels of TNF-α, IL-6, and IL-1β remained unaffected	Exercise training ↓ pro-inflammatory effects, ↓ catabolic wasting process associated with the progression of CHF
Mattusch et al., 2000 (37)	14 male runners (25–40 years, mean 34 years) for a marathon training, and 11 (31–52 years) non-exercising males	9 months’ intense marathon training: 50 min × 3–4 sessions × 75% VO2peak (∼anaerobic threshold)	Exercise training ↑ mean distance run per week, ↑ running velocity during a maximal treadmill test, ↓ base-line CRP	Intensive regular exercise training has a systemic anti-inflammatory effect
Tisi et al., 1997 (53)	67 IC patients, randomized to exercise (n = 22), control (n = 17), and PTA (n = 28) group	4 weeks’ active and passive leg exercise (1 hour daily), plus 1.61 km (1 mile) walking	Claudicants ↑ urinary ACR following treadmill exercise test. Exercise training ↓ SAA at 6 months, ↓ CRP at 3 months, and progressively ↓ the post-exercise increase in ACR	Exercise training is associated with symptomatic improvement and ↓ inflammatory markers
Resistance
Donges et al., 2010 (43)	102 sedentary subjects, assigned to resistance (n = 35), aerobic (n = 41), or control (n = 26) group	10 weeks’ resistance (10 repetitions, ∼75% 1RM) or aerobic (70%–75% HRmax, 30–50 min) exercise training	Resistance and aerobic exercise training ↓ 32.8% and 16.1% CRP, respectively. Aerobic group ↑ all aerobic fitness measures and ↓ IA-FM and ↓ body mass. Resistance group ↓ TB-FM, ↑ 46.3% upper and ↑ 56.6% lower body strength	Only resistance exercise training significantly ↓ CRP (aerobic exercise training is less effective).
Ogawa et al., 2010 (36)	21 elderly women (85.0 ± 4.5 years) were volunteered to participate in exercise training	12 weeks’ low-intensity resistance exercise training: 40 min × ≧ 1 per week, 1 set of 10 repetitions for the foot press, front traction, and shoulder press; 2 sets of 10 repetitions for vertical traction	Training ↓ CRP, TNF-α, SAA, HSP70, IGF-I, and insulin. And training-induced ↓ CRP and TNF-α are associated with ↑ muscle thickness	Resistance training may assist in maintaining or improving muscle volume and ↓ low-grade inflammation
Interval
Byrkjeland et al., 2011 (63)	80 stable CHF, randomized to exercise (n = 40) or control (n = 40) group	4 months’ high-intensity group-based aerobic training: 3 intervals of high-intensity (RPE 15–18) and 2 moderate (RPE 11–13) exercise, 50 min × 2 per week	Only IDCM patients ↓ CRP, ICAM-1, TGF-β, and TNF-α after training	Anti-inflammatory effects of exercise only applied in IDCM patients. The etiology of CHF affected the inflammatory profile and the effect of exercise training
Lamina et al., 2009 (54)	43 older male HTN patients with erectile dysfunction, randomized to exercise training (n = 22) or control (n = 21) group	8 weeks’ interval exercise training (60%–79% HRR), 45 to 60 min/day	Exercise training ↑ erectile function and ↓ CRP	Exercise program is a possible effective non-invasive and non-pharmacologic management of erectile dysfunction in male HTN patients
Nawaz et al., 2001 (58)	52 stable IC, randomized to supervised upper (n = 26) or lower (n = 26) limb exercise group; and a parallel control group (n = 15)	6 week arm cranking or leg ergometer interval training: 20 min × 2 times per week × 70%–80% HRmax intensity.	An acute bout of lower limb exercise ↑ neutrophil CD11b and CD66b expression. Both upper and lower limb exercise training ↑ walking distance without affecting acute bout exercise-induced neutrophil CD11b, CD66b and circulating vWFs	Upper limb exercise training may offer certain advantages over lower limb programs by improving cardiorespiratory function and walking capacity, avoiding the potentially deleterious systemic inflammatory responses

↑ = increase; ↓ = decrease; AAA = abdominal aortic aneurysm; ACR = albumin-creatinine ratio; ALP = alkaline phosphatase; AMI = acute myocardial infarction; BNP = brain natriuretic peptide; CAD = coronary artery disease; CBF = calf blood flow; cfPWV = carotid-femoral pulse-wave velocity; CHF = chronic heart failure; CK = creatine kinase; CRP = C-reactive protein; CVD = cardiovascular disease; EC-SOD = extracellular superoxide dismutase; EF = ejection fraction; FGF-2 = fibroblast growth factor-2; FMD = flow-mediated dilation; GXT = graded exercise test; HF = heart failure; HDL = high-density lipoprotein; HMGB1 = high-mobility group box-1; HOMA-IR = homeostasis model assessment of insulin resistance; HRR = heart rate reserve; HTN = hypertension; IA-FM = intra-abdominal fat mass; IC = intermittent claudication; ICAM-1 = intercellular adhesion molecule-1; IDCM = idiopathic dilated cardiomyopathy; IGF-1 = insulin-like growth factor-1; IL-6 = interleukin-6; IMT = carotid intima-media thickness; iNOS = inducible nitric oxide synthase; IR = insulin resistance; LDL = low-density lipoprotein; LVEF = left ventricular ejection fraction; LVEDV = left ventricular end-diastolic volume; Mac-1 = macrophage-1 antigen; MCP-1 = monocyte chemoattractant protein-1; MET = metabolic equivalent task; Mig = monokine induced by interferon-gamma; MS = metabolic syndrome; NF-κB = nuclear factor-kappa B; NYHA = New York Heart Association; PAD = peripheral arterial disease; PCI = percutaneous coronary intervention; PTA = percutaneous transluminal angioplasty; QoL = quality of life; RM = repetitive maximum; RPE = rating of perceived exertion; SAA = serum amyloid A protein; SBP = systolic blood pressure; TBARS = thiobarbituric acid reactive substances; TB-FM = total body fat mass; TB-LM = total body lean mass; TGF-β = transforming growth factor-beta; TNF-α = tumor necrosis factor-alpha; TNF-αSR1 = soluble tumor necrosis factor-alpha receptor 1; VCAM-1 = vascular cell adhesion molecule-1; VEGF-A = vascular endothelial growth factor-A; VLA-4 = very late antigen-4; VO₂max = maximal oxygen consumption; WIQ = walking impairment questionnaire.

Acute exposure exercise studies

Eight acute-bout studies have reported the effects of two modes of exercise (aerobic, with average duration of 40 min (28–33); and resistance, with average 2–3 sets × 12–20 repetitive maximum (34,35)) on cytokines in plasma and skeletal muscle biopsies. In the aerobic studies, a graded treadmill test protocol (29,33) or high-intensity activity (75 percentage of maximal oxygen consumption (% VO₂max)) (28,30) elicited significant increases in IL-6, TNF-α, and CRP post-exercise, and this increment of pro-inflammatory cytokines was augmented in coronary artery disease patients compared to healthy controls (33). By contrast, Markovitch et al. reported that moderate physical activity (50% VO₂max) (31) had neither pro-inflammatory (IL-6 and CRP) nor anti-inflammatory (IL-10) effects. However, another two moderate aerobic studies revealed increased IL-6 (28,32) and IL-1 receptor agonist (IL-1ra) (32) after exercise. In addition, acute intense resistance exercise-induced micro-damages of skeletal muscle leading to pro-inflammatory cytokines (e.g. IL-6, IL-8, monocyte chemoattractant protein-1 (MCP-1), CRP) production (34,35).

Despite the diverse results of the effects of an acute bout of exercise on inflammatory processes, one could safely conclude that strenuous exercise is required for the activation of pro- inflammatory pathways, while moderate activity does not seem to affect inflammation. Furthermore, the inconsistent results of acute bouts of exercise suggest the need for further studies with definite intervention intensities, larger sample sizes, well-controlled subject recruitment, and precise sample collection time points.

Chronic exposure (exercise training) studies

Twenty-six randomized control studies, one uncontrolled trial (36), and one non-randomized control trial (37) were reviewed. Most investigators incorporated aerobic exercise. In two studies resistance training was utilized. Aerobic exercise activities included home-based (38,39)/treadmill walking (40), cycling, arm cranking, marathon running, hiking, exercise-based cardiac rehabilitation programs, group exercise, and yoga practice. The training frequency and duration ranged from 3–6 times per week, 3 weeks, to 2-year periods in total. Most studies used moderate-intensity exercise (50%–74% VO₂max).

Moderate aerobic exercise training not only significantly increased excise capacity in a dose-dependent manner (41), improved traditional cardiovascular risk factors (i.e. lipids, glucose tolerance, and visceral fat) (42), but also decreased oxidative stress and reduced pro-inflammatory mediators provoked by acute intense exercise in both healthy subjects (42–45) and cardiovascular patients (46–53). Endothelial function (nitric oxide synthesis and vasodilatation) (54–57), leukocyte activating components (e.g. adhesion molecules: Mac-1, VLA-4, CD66; or inflammatory protein: high-mobility group box-1 (HMGB-1)) (58,59), autonomic regulation (60), and cardiac remodeling (60) may also benefit from moderate endurance training. These positive results seem to improve QoL further and enhance event-free survival rate (55,61). In addition, the beneficial effects of physical training are evident in all age groups: children (57), young individuals (62), and middle-aged to elderly populations (36,54). Similarly, moderate-to-high-intensity exercise training tends to have anti-inflammatory properties through suppressing pro-inflammatory cytokines secretion (54,63) and improving endothelial anti-inflammatory function (54). Based on previous references (i.e. studies discussing the effects of exercise training on inflammatory processes), the weekly exercise training volume for cardiovascular patients (including post-infarction, coronary artery disease, and heart failure patients) to reduce pro-inflammatory factors or to increase anti-inflammatory factors is in the range from 4 to 9 MET-hours (39,46,49,55,60). However, for healthy subjects (44) or people who have cardiovascular risks (e.g. moderate hypertension (41) or smoking (45)), a weekly exercise training volume at the level of 8–27 MET-hours was not sufficient to provide beneficial inflammatory profiles. Czepluch et al. (62) showed that strenuous exercise training volume (64.2–75.6 MET-hours per week) can reduce monocyte chemotaxis in healthy young subjects. Therefore, the lowest amount of weekly exercise training volume to reach inflammatory benefits may need to be individualized.

Resistance programs seem also to possess anti-inflammatory features (36). Indeed, moderate-intensity resistance exercise has superior impact on anti-inflammatory course compared to moderate aerobic training (43). On the contrary, however, strenuous either resistant or aerobic physical training may have detrimental effects including inhibition of proper cardiovascular repair through declining monocyte chemotaxis (TGF-β and MCP-1) and angiogenesis (VEGF-A). Interestingly these effects persist for a month without continuing exercise (62). Further longitudinal observations are required to elucidate whether this is a sound adaptation to increased stimuli or an unfavorable reaction to overtraining. Hence, although regular and moderate physical exercise is usually quoted to enhance the resistance against inflammation, intense exertion probably has adverse consequences (62).

In summary, acute strenuous exercise seems to activate inflammatory pathways. Plasma IL-6 increases concomitantly and up to 100-fold of the basal level from contracting skeletal muscle (3,64). The IL-6 increment stimulates AMP-activated protein kinase (AMPK) and acetyl CoA carboxylase phosphorylation, therefore enhances lipid turnover, stimulating lipolysis as well as fat oxidation. Elevated IL-6 may subsequently lead to CRP elevation (as an acute reactance from liver) that can be more evident during late recovery of exercise. In response to long-term physical training, regular exercise leads to an enhancement of glycogen synthase, which will consequently augment muscle glycogen storage and fat β-oxidation. Under the high muscle glycogen environment, fat oxidization becomes the major substrate during exercise through the enhancement of muscular IL-6 receptor (IL-6R) expression, thus transcription rate of circulating IL-6 is slower in a trained versus an untrained subject (Table II) (3,64).

Table II. Summary of the exercise effects on inflammatory/anti-inflammatory mediators.

Inflammatory/anti-inflammatory markers	Physical exercise
	Acute
	Moderate	Heavy	Regular
Pro-inflammatory cytokines
IL-6	↔ or ↑	↑↑	↓↓
TNF-α	↔	↑	↓
CRP	↔ or ↑	↑	↓
IL-1β	↔	↑	↓
MCP-1	↔	↑	↓
IL-8	↔	↑	↓
TNF-αSR-1 or -2	↔	↑	↓
Anti-inflammatory cytokines
IL-1rα	↑	↑↑	↑
IL-10	↔ or ↑	↑	↑
Adiponectin	↔	↑ or ↓	↔ or ↑

↑ = increase; ↑↑ = marked increase; ↓ = decrease; ↓↓ = marked decrease; ↔ = no difference; CRP = C-reactive protein; IL = interleukin; MCP = monocyte chemoattractant protein; TNF = tumor necrosis factor; TNF-αSR = soluble tumor necrosis factor-α receptor.

Exercise and thrombogenesis

An acute bout of vigorous exercise has been shown to increase the risk of primary cardiac arrest, whereas regular moderate physical training contributes to the reduction of thrombotic cardiovascular events (65–67). A number of hemostatic changes occur with acute and chronic exercise regimens involving platelet function/activity, coagulation factors, fibrin formation, fibrinolysis, and blood rheology. Strenuous exercise-induced transient hypercoagulable state is associated with shortened activated partial thromboplastin time (aPTT), increased thrombin generation (68), elevated plasma levels of coagulation factor VIII (FVIII) and von Willebrand factor (vWF) (69,70), as well as an increase in platelet count (71,72) and reactivity (65,72–74). These reversible changes are more pronounced in sedentary than physically active subjects (65), but even athletes are not excluded from the risks (75). In contrast, an exercise conditioning program results in a down-regulated hemostatic potential by reducing platelet activation (73,76–78), diminishing procoagulant levels, and augmenting fibrinolysis activity (79,80), therefore contributing to a reduced risk of thrombotic events (Table III) (81).

Table III. Studies of exercise effects on thrombogenesis.

Source	Subjects	Exercise	Main results	Summary
Acute bout exercise
Aerobic
Menzel et al., 2011 (92)	25 young, healthy, untrained subjects	80% (moderate) and 100% (strenuous) of individual AT, 60 min/session	Strenuous exercise ↑↑ coagulation activity and fibrinolysis more than moderate exercise	Moderate exercise ↑ fibrinolytic activity. Strenuous exercise ↑↑ coagulation ability and profound fibrinolytic function
Sossdorf et al., 2011 (81)	16 healthy young male: untrained (n = 8, VO2max < 50 mL/kg/min) and trained (n = 8, VO2max > 65 mL/kg/min) group	A single bout of bicycle ergometer: 80% AT, 90 min	Exercise ↑ MP numbers and procoagulant activities. Trained individuals had a lower number TF^+ monocyte-derived MP and faster recovery than untrained individuals	Exercise ↑ MP and ↑ exercise-induced hemostatic potential in untrained individuals
Chen et al., 2010 (68)	24 healthy young men	A graded cycle ergometer exercise test (up to VO2max)	Low and high shear stresses ↑ PDMP and TG peak height and rate. Strenuous exercise ↑ FV/Va- /FVIII- /TF-rich and PS-exposing PDMP counts, ↑ PDMP procoagulant activity, and returned to baseline 2 h afterward	Strenuous exercise modestly contributes to the acceleration of shear-induced TG by ↑ release of procoagulant PDMP
Tripette et al., 2010 (139)	11 sickle cell trait carriers; 11 subjects with normal hemoglobin	Submaximal exercise, 15 min	↑ Erythrocyte aggregation and blood viscosity	Acute exercise-induced erythrocyte aggregation and blood viscosity did not accompany elevated oxidative stress. It may be relevant to increased plasma fibrinogen or vWF levels
Menzel et al., 2009 (94)	24 young and 24 middle-aged healthy non-smokers	Moderate exercise test: 80% AT, 60 min	Middle-aged group ↑ FII, VII, VIII, IX, XI, XII, F1 + 2, tPA:Ag, tPA:Ac, plasminogen. Exercise induced similar increment in both groups, but middle-aged subjects ↑↑ F1 + 2 and TAT after exercise test	Blood coagulation and fibrinolysis have an age-related rise, which moderate exercise both ↑ coagulation and fibrinolysis in middle age
Petidis et al., 2008 (85)	60 subjects: healthy volunteers (n = 18), essential HTN (n = 26), stable CAD patients at least 3 months (n = 16)	A treadmill exercise test: modified Bruce protocol, 15 min	↑ thromboxane B2, ↑ endothelium prostacyclin release, ↓ platelet aggregation in all groups, significantly more in healthy volunteers than in CAD patients (HTN lying in between)	Exercise test ↑ activation of sympathetic nervous system, ↑ TXB2, but ↓ platelet aggregation, possibly to counterbalance the prothrombotic state. In endothelium dysfunction groups (HTN and CAD), the reduced platelet aggregation is attenuated
Kobusiak-Prokopowicz et al., 2006 (89)	35 CAD on aspirin (75 mg daily) treatment patients; 10 age- and gender-adjusted healthy controls	Treadmill testing: Bruce protocol	Exercise ↑↑ platelet P-selectin (without stimulation) in patients than in the control group	Platelet P-selectin ↑ in CAD patients and ↑↑ after exercise test
Kratz et al., 2006 (87)	32 healthy participants	26.2 mile (42.2 km) marathon	↑ hematocrit, platelet count, leukocytosis, red blood cell fragments, and platelet clumps	Marathon exercise ↑ in vivo inflammation, tissue damage, hemolysis, and platelet activation
Womack et al., 2006 (97)	15 males	Cycle ergometer tests: above (77% VO2peak) or below (41% VO2peak) lactate threshold (LT)	Exercise ↑ tPA:Ag, ↓ PAI-1:Ac, and the changes are greater in intensity above LT. Only exercise intensity above LT ↑ tPA:Ac	Exercise intensity above LT elicits greater fibrinolytic responses than exercise below LT
Wang, 2004 (91)	18 healthy men	Intense exercise of cycle ergometer: 80% VO2max, 40 min	↑ vWF:Ag, vWF:Ac; ↑ shear- or ristocetin-induced platelet aggregation; ↑ vWF binding to platelets and vWF-mediated GP IIb/IIIa activation at high shear flow	Intense exercise ↑ shear stress-induced platelet aggregation by ↑ vWF binding to platelets and the subsequent activation of GP IIb/IIIa complexes
Burns et al., 2003 (93)	Patients with IC (n = 10); age- and sex-matched smokers ([S], n = 5); non-smokers ([NS], n = 5) without peripheral vascular disease	Treadmill exercise test	IC group ↑ TAT, F1 + 2 after exercise than S or NS group; Pre and post-exercise, fibrin turnover in IC was similar to S controls, but higher than NS controls	Exercise test ↑ excessive production of thrombin in IC. IC may have prothrombotic state by defective fibrinolysis
Hilberg et al., 2003 (95)	16 healthy men	Treadmill ergometer test at 90% individual AT for 60–120 min	↑ F1 + 2, TAT, TTP; ↑↑ PAP, D-dimer, tPA:Ag and Ac; ↓ PAI-1:Ac	Endurance exercise generates a favorable condition for fibrinolysis than blood coagulation in healthy young men
Hilberg et al., 2003 (96)	15 healthy men	Cycle ergometer maximal test with durations of 15, 45, and 90 s	↑ F1 + 2 and TTP; ↑↑ PAP, tPA:Ag	Maximal short-term exercise does not lead to a relevant activation of blood coagulation in healthy men, but fibrinolysis is clearly activated
Ivey et al., 2003 (98)	18 untrained stroke patients with chronic hemiparetic gait deficit	Treadmill walking: 60% HRR (low-to-moderate intensity), 20 min	Exercise ↑ tPA:Ac and ↓ PAI-1; but 1 h after exercise, tPA activity still ↑, PAI-1:Ac ↓ than baseline	Moderate aerobic walking exercise can improve fibrinolysis profiles in chronic stroke patients
Nageswari et al., 2000 (137)	20 subjects	Mild exercise, 15 min	↑ whole blood viscosity, and it is negatively proportional to shear stress levels; ↑ red cell rigidity; ↔ red cell aggregation and hematocrit; ↑ basal perfusion; ↓ reactive hyperemia perfusion index	Short-term mild exercise significantly alters hemorheological and microcirculatory parameters
Placanica et al., 1999 (90)	20 patients with previous MI; 40 patients with exercise-induced angina; 20 healthy controls	Bicycle exercise test	Ischemic heart disease patients ↑ PF-4. Exercise ↑ PF-4 in patient groups, but not in healthy controls or those patients without angina or ST-depression	Exercise ↑ PF-4 is associated with clinically significant coronary artery disease
Weiss et al., 1998 (70)	12 young men	Treadmill for 1 hour: Moderate (82% HRmax or 68% VO2max). Very heavy (94% HRmax or 83% VO2max)	Moderate exercise ↑ tPA:Ag and PAP; ↔ F1 + 2, TAT and fibrinopeptide A. Very heavy exercise ↑↑ tPA and PAP (exceed normal values), ↑ F1 + 2, TAT fibrinopeptide A (within normal values)	Moderate exercise results in increased plasmin formation only, while at very heavy exercise generation of plasmin seems to exceed that of thrombin and fibrin
Kestin et al., 1993 (65)	12 physically active and 12 sedentary individuals	Standardized treadmill exercise test	Exercise test ↑ thrombin-induced platelet GP Ib and IV expression, but not in active group	Strenuous exercise ↑ platelet activation and hyperreactivity in sedentary subjects but not physically active subjects
Resistance
Nagelkirk et al., 2010 (102)	Healthy young women	6 sets of 10 leg extension repetitions at 70% of 1RM	↑ tPA:Ac; ↔ TAT and PAI-1:Ac; %body fat is negatively correlated to tPA response to exercise and positively related to PAI-1	Acute resistance exercise ↑ tPA:Ac, but does not change TAT and PAI-1:Ac. Women with elevated body fat may be at increased risk of thrombosis
deJong et al., 2006 (83)	14 low-to-moderate risk men with CAD	1 set of 10 repetitions to volitional fatigue	↔ vWF; ↑ tPA immediately after exercise; ↓ PAI:Ac immediately after exercise and remained ↓ at 1 h post-exercise	An acute bout of resistance exercise improves fibrinolytic potential in men with CAD without elevating thrombotic potential
Ahmadizad et al., 2003 (86)	13 healthy men	3 different intensities of resistance exercise: 40%, 60%, and 80% of 1RM	Resistance exercise ↑ platelet count, PCT, and MPV, but not in intensity-dependent manner. Only high-intensity (80% 1RM) resistance exercise ↑ platelet aggregation, ADP, and β-TG	Resistance exercise ↑ platelet, PCT, and MPV. High-intensity exercise ↑ platelet aggregation and β-TG
Yalcin et al., 2003 (138)	10 healthy young men	Wingate (heavy) anaerobic power test	↑ Lactate 10-fold after exercise; ↑ erythrocytes counts immediately after exercise; ↑ WBC (granulocyte numbers and activity); ↓ erythrocyte deformability and aggregation	A single bout of heavy exercise may induce significant hemorheological deteriorations and last for 12 h
Regular exercise or training
Aerobic
Jahangard et al., 2009 (108)	20 postmenopausal sedentary women, randomized to training (n = 10) and control (n = 10) groups	3 weeks’ submaximal aerobic cycling: 35 min × 10 sessions (5 min warm-up, 25 min aerobic training with 70% HRmax, 5 min active and 15 min passive recovery), 3 times/week	Exercise training ↓ fibrinogen, vWF-Ag, PAI-1:Ac, PAI-1:Ag; ↑ PT, PTT, tPA:Ac and tPA:Ag	3 weeks’ submaximal aerobic exercise ↑ fibrinolytic activity on postmenopausal women
Wang, 2005 (91)	30 healthy young men, randomized to training (n = 15) or control (n = 15) groups	8 weeks’ bicycle ergometer training: 60% VO2max, 30 min, 5 days/week, and then were deconditioned for 8 weeks	Exercise training ↓ strenuous exercise-evoked ASIPA through ↑ platelet-vWF binding and P-selectin expression. Deconditioning reverses the effects of training on resting and post-exercise state	Exercise training ↓ resting and severe exercise-promoted ASIPA by ↓ vWF binding and P-selectin expression on platelets. The training effects are reversed after deconditioning
Wang et al., 1997 (78)	16 healthy sedentary women, randomized to control (n = 8) and exercise (n = 8) groups	2 consecutive menstrual cycle exercise training: 50% VO2max × 30 min × 5 days and then were deconditioned for 3 menstrual cycles	Exercise training ↓ resting platelet function and acute exercise-potentiated platelet activation, ↑ plasma nitrite and nitrate and platelet cGMP; deconditioning reversed training effects	Training-induced platelet functional changes in women in the midfollicular phase may be mediated by nitric oxide
El-Sayed et al., 1995 (79)	25 young subjects, randomized to exercise (n = 13) or control (n = 12) group	12 weeks’ exercise training: 30 min, 3 times per week, at 70% HRmax (6 weeks) and 80% HRmax (6 weeks)	Maximal exercise ↑ aPTT, thrombin clotting time, FVIII:Ag, FVIII:Ac, plasminogen in both groups, no differences between groups	Maximal exercise transiently ↑ blood coagulation and activates blood fibrinolytic activity, rather than at rest. Moderate exercise training did not change coagulation or fibrinolysis
Wang et al., 1995 (67)	23 healthy young men, randomized to exercise training (n = 11) and control (n = 12) groups	8 weeks’ exercise training: 60% VO2max, 30 min/day, 5 days/wk, followed by 12 weeks’ decondition	Exercise training ↓ resting and acute exercise-provoked platelet adhesiveness and aggregability; training effects were reversed after deconditioning	Exercise training ↓ platelet adhesiveness and aggregability but are reversed back to the pre-training state after deconditioning
Ponjee et al., 1993 (107)	20 sedentary males; 15 sedentary females	9 months’ exercise training: 3–4 times/week with increasing intensity. After 24 and 36 weeks all subjects ran a 15 km and a half-marathon (21 km) race	↔ plasma FVIII, vWF, fibrinogen, TAT. Fibrinogen ↑ after 21 km race and was still ↑ 5 days later	Physical stress ↑ plasma fibrinogen. Regular physical fitness program does not induce a long-term activation of the hemostatic system
Resistance
Clark et al., 2011 (109)	16 young adults	Resistance training, 3 days/wk, 4 weeks. Low-load resistance exercise with blood flow restriction (BFRE), 30% of strength. High-load resistance exercise (HLE), 80% of strength	Strength ↑ 8% with BFRE; ↑ 13% with HLE. No changes in fibrinogen, D-dimer, CRP	Both BFRE and HLE ↑ strength without alternating nerve or vascular function. Both single bout protocols ↑ fibrinolytic activity without altering coagulation or inflammation
Interval
Bye et al., 2009 (110)	11 metabolic syndrome patients	16 weeks AIT: 4 × 4 min intervals at 90% HRmax, with 3-min active recovery at 70% HRmax between intervals, total exercise time 40 min	↑ 11 steroid hormone-mediated signaling; ↓ 7 blood clotting, cell adhesion, and steroid metabolism process; ↓ protein levels of arginase 1 and vWF	AIT improves endothelial function, and cardiovascular risk profile of metabolic syndrome patients

↑ = increase; ↓ = decrease; ↑↑ = marked increase; Ac = activity; ADP = adenosine diphosphate; AF = atrial fibrillation; Ag = antigen; AIT = aerobic interval training; ASIPA = alternating shear-induced platelet aggregation; AT = anaerobic threshold; β-TG = β-thromboglobulin; cGMP = guanosine 3’,5’-cyclic monophosphate; F1 + 2 = prothrombin fragment 1 + 2; FVIII = coagulation factor VIII; GP = glycoprotein; Hb = hemoglobin; IC = intermittent claudication; LT = lactate threshold; MP = microparticle; MPV = mean platelet volume; PA = platelet aggregation; PAI = plasminogen activator inhibitor; PAP = plasmin-alpha 2 antiplasmin complex; PCT = plateletcrit; PDMP = platelet-derived microparticle; PF-4 = platelet factor-4; PS = phosphatidylserine; PT = prothrombin time; PTT = partial thromboplastin time; TAT = thrombin-antithrombin complex; TF = tissue factor; TG = thrombin generation; tPA = tissue plasminogen activator; TTP = total thrombin potential; TXB2 = thromboxane B2; vs. = versus; vWF = von Willebrand factor.

In contrast to traditional aerobic exercise prescriptions, resistance training has become an increasingly important component of cardiac rehabilitation (82). When appropriately prescribed and supervised, it has favorable effects on bone mineral density, body composition, muscular strength and endurance, glucose metabolism, selected coronary risk factors, and other health-related variables. Moreover, resistance exercise has also been shown to attenuate the rate-pressure product through potentially decreasing cardiac demands during daily carrying or lifting activities (83). However, acute cardiac events due to an occlusive thrombus often occur during or after heavy physical exertion, and concern of thrombosis has been suggested due to significant vascular compression and release of several contributing mediators (e.g. catecholamines or bradykinin) (84). Hence before resistance training is widely recommended, the responses to acute and/or chronic resistance exercise in regard to thrombogenetic risks of patients should be evaluated thoroughly.

Acute-exposure exercise studies

It has long been known that strenuous exercise induces a hypercoagulable state in circulation. Exercise bouts of varied intensity and mode have induced significant levels of differences in platelet reactivity, hemostasis, and fibrinolysis. Intense physical exercise-induced sympathetic nerve activation and stress hormone release (85) usually accompany peripheral blood thrombocytosis (86,87) and platelet hyperreactivity (65,85,88,89). Elevated catecholamine levels not only increase platelet surface α₂-adrenoreceptor numbers, augment their binding affinity to fibrinogens, enhance β-thromboglobulin (β-TG) and platelet factor-4 (PF-4) discharge (90), but also amplify platelet P-selectin expression and glycoprotein IIb/IIIa activation (91), consequently leading to platelet aggregation and accelerated thrombus formation. However, unlike intensive exercise, acute moderate programs do not elicit platelet activation (88).

In addition, although vigorous exercise augments both blood coagulation and fibrinolysis, the profounder increments of plasma vWF (91), FVIII (92), thrombin (68,92–94), and procoagulant microparticles (68,81) shift the net effect toward a hypercoagulable state (95,96). In contrast, moderate physical activity promotes a fibrinolytic state through increasing tissue plasminogen activator (tPA) and decreasing plasminogen activator inhibitor-1 (PAI-1) without provoking coagulable mediators (70,97,98). In view of the possible links between acute exercise and thrombotic consequences, previous studies speculated that intense physical activity may trigger acute coronary thrombosis and occlusions in sedentary individuals or those with pre-existing cardiovascular disease (99) by disrupting the delicate balance between coagulation and fibrinolysis, as well as up-regulating platelet activity (65,85,90,100).

Nevertheless, blood coagulation and fibrinolysis have an age-related rise, and moderate exercise increases both coagulation and fibrinolysis in middle-aged individuals (94). Furthermore, acute-bout strenuous exercise-induced pathological high shear stress may enhance the influx of extracellular Ca²⁺ and the binding of vWF to platelet glycoprotein Ib (GP Ib), thus increasing the generation of microparticles and platelet clumps (68,81,91). Although a strenuous exercise-induced prothrombotic state (e.g. increased procoagulant microparticle release or thrombin formation (68) or leukocytosis (87)) can be augmented by CVD risk factors (e.g. smoking (93) or sedentary lifestyle (65,81)), the offset effects to counterbalance platelet aggregation/ activation and thrombotic system in healthy subjects are attenuated in patients with hypertension, CVD, or intermittent claudication (85,89,93).

Besides, both acute low-intensity and high-intensity resistance exercise may increase circulating tPA levels through stimulating endothelium tPA release (vascular compression) or reducing hepatic tPA clearance (perhaps by decreased liver blood flow) (83,101). However, thrombotic parameters (e.g. TAT, F1 + 2, or PAI-1) remained unchanged after intense resistance exercise (86,102).

Therefore, the acute hemostatic responses to aerobic exercise follow in an intensity-dependent manner, but the responses to resistance exercise are more dependent on exercise ‘duration’ or ‘volume’. Overall, is this beneficial fibrinolytic profile transitory or maintainable after resistance exercise? Also, whether these indirect biomarkers can actually represent clinical manifestations of thrombosis during and/or after exercise is still uncertain.

Chronic exposure (exercise training) studies

Regular exercise has been proved to reduce cardiovascular morbidity and mortality through modifying traditional and novel risk factors (103). Moderate exercise can effectively ameliorate platelet adhesiveness and aggregability by down-regulating adhesion molecule activation and expression, decreasing α₂-adrenergic receptor performance, and reducing vWF-platelet interaction (67,78,104). Moreover, long-term training up-regulates platelet and endothelial nitric oxide (NO) release and bioavailability, enhances antioxidants levels (e.g. superoxide dismutase), and hence reduces the susceptibility to oxidative stress and lipid peroxidation (78,105). Suppression of platelet reactivity induced by physical training may in turn diminish risks of thrombotic events.

Studies assessing the effects of exercise on blood coagulation and fibrinolysis are limited. A cross-sectional study showed no significant difference in prothrombin time (PT) and aPTT at rest and post-exercise among sedentary subjects, joggers, and marathoners (106). These results were confirmed by a randomized control trial (79). Additionally, aerobic training programs did not change resting levels of FVIII coagulant activity in sedentary individuals either (107). However, one recent study showed that short-term (3 weeks) moderate aerobic training can effectively decrease coagulation function (i.e. training-induced fibrinogen, vWF:Ag, PAI-1:Ag, and PAI-1:Ac suppression) and increase fibrinolytic function (i.e. training-induced PT, aPTT, tPA:Ag, and tPA:Ac elevation) in sedentary postmenopausal women (108).

Similarly, low- or high-load resistance exercise training not only improves muscle strength in a short period (4 weeks), but increases fibrinolytic advantages without evoking coagulation or inflammatory responses (145). Furthermore, moderate-to-high aerobic interval training (AIT; 4 × 4 90% HRmax and 70% HRmax periodic program) had shown the ability to reduce blood clotting gene transcription in metabolic syndrome patients (110). Therefore, various exercise training protocols can be utilized to reduce thrombus formation and enhance clot degradation, perhaps through increasing plasmin formation and activity (Table IV) (66).

Table IV. Summary of the effects of exercise on markers of thrombosis/fibrinolysis.

Thrombogenic/fibrinolytic markers	Physical exercise
	Acute
	Moderate	Heavy	Regular
Platelet function	↔	↑	↓
α2-Adrenoreceptor expression	↔	↑	↓
Fibrinogen binding	↔	↑	↓
β-TG	↔	↑	↓
PF-4	↔	↑	↓
P-selectin expression	↔	↑	↓
GP IIb/IIIa activation	↔	↑	↓
Coagulation	↔	↑	↔ or ↓
vWF	↔	↑	↔ or ↓
FVIII	↔	↑	↔ or ↓
TAT	↔	↑	↔ or ↓
F1 + 2	↔	↑	↔ or ↓
TG	↔	↑	↓
Microparticles	↔	↑	↓
Fibrinolysis	↔ or ↑	↑↑	↑
tPA	↔ or ↑	↑↑	↑
PAI-1	↔ or ↓	↓	↓

↑ = increase; ↑↑ = marked increase; ↓ = decrease; ↓↓ = marked decrease; ↔ no difference; β-TG = beta-thromboglobulin; GP = glycoprotein; FVIII = coagulation factor VIII; F1 + 2 = prothrombin fragment 1 + 2; PAI-1 = plasminogen activator inhibitor-1; PF-4 = platelet factor-4; TAT = thrombin-antithrombin complex; TG = thrombin generation; tPA = tissue plasminogen activator; vWF = von Willebrand factor.

Our understanding of the interindividual variability in response to exercise, the direct effects of various exercise on thrombosis, and the plausible mechanisms underlying these effects is incomplete and fragmented. As data on the effects of exercise on blood coagulation and fibrinolysis are limited, further longitudinal investigations in regard to exercise responses are required.

Mechanisms of exercise-induced inflammatory processes and the implications for thrombogenesis

Growing evidence has indicated that atherosclerosis (6) and CVD (111) are intimately linked to chronic vascular inflammation. There are also considerable data suggesting that inflammatory processes have fundamental roles in thrombogenesis by promoting coagulation, platelet reactivity, and suppressing fibrinolysis. Inflammatory cells, adhesion molecules, cytokines, and procoagulant microparticles appear to be associated with the process (18). An acute inflammatory response triggers thrombus formation by enhancing platelet activation (up-regulated sCD40L and P-selectin expression), promoting endothelial damage and erythrocyte aggregability/rigidity (reduced NO bioavailability), inducing leukocyte activation and monocyte TF expression (12), and increasing lipid peroxidation (Figure 1) (111). Several inflammatory biomarkers (e.g. IL-6, CRP, adiponectin) are associated with prothrombotic stimulation.

Figure 1. Pathogenesis of vascular thrombosis: from physical inactivity, inflammation, to thrombogenesis. ↑ = increased; ↓ = decreased; EMP = endothelial-derived microparticles; NO = nitric oxide; PAI-1 = plasminogen activator inhibitor-1; PMP = platelet-derived microparticles; sCD40L = soluble CD40 ligands; tPA = tissue-type plasminogen activator; vWF = von Willebrand factor.

Moreover, physical inactivity has been considered as an independent risk factor of CVD. The interrelationship between physical activity levels, inflammatory processes, and thrombogenic pathways promoted research in the area of exercise-related health benefits. However, exercise interventions mostly focused on either inflammation (Table I) or thrombogenesis (Table III), and studies focusing on both aspects of atherothrombosis are still sparse.

Understanding the basic biological mechanism of exercise-induced inflammation and its implication in thrombogenesis is important, as novel diagnostic/prognostic markers or therapeutic targets might emerge. Basal IL-6 and CRP levels are lower in physically active subjects (27), but whether circulating levels of IL-6 and CRP are more or less superior to other inflammatory markers associated with obesity (e.g. adiponectin) and a sedentary lifestyle (e.g. lipid metabolism) has not been established (113). In addition, the impact of exercise on chronic inflammation-induced endothelial dysfunction and erythrocyte rheology is also unclear.

Inflammation versus thrombogenesis: microparticle (MP)

Chaar et al. reported that strenuous ramp cycling increased plasma IL-6 and platelet- (PMP) and leukocyte-derived microparticle (LMP) levels, but did not change other microparticles derived from endothelium or erythrocyte, and soluble adhesion molecules. Hence strenuous exercise-induced inflammation-activated release of procoagulant microparticles may be independent of endothelial activation (114). However, Mesri et al. demonstrated that polymorphonuclear cell (PMN)-derived microparticles are closely linked to endothelial activation and IL-6 and IL-8 secretion (115). Therefore the results suggested that circulating microparticles may be considered as biomarkers and potential bio-effectors of the inflammatory and thrombotic responses induced by physical exercise. In other words, exercise-induced microparticle amounts, subtypes, and characteristics can be surrogates to reflect the severity of vascular inflammation and thrombosis.

Inflammation versus thrombogenesis: soluble CD40 ligand (sCD40L)

Soluble CD40L is a biomarker of platelet activation, which is secreted by platelet granules. Soluble CD40L may trigger inflammatory mediators’ up-regulation, leading to matrix metalloproteinase production and subsequently to the activation of the coagulation cascade. On the one hand, acute moderate exercise (cycling or jogging) may reduce plasma sCD40L levels without changing sCD40 levels in well-trained athletes (116). On the other hand, 20 weeks of aerobic training would not only improve exercise capacity and QoL, but reduce sCD40L and P-selectin expression without inducing proinflammatory cytokines in mild to moderate chronic heart failure patients (117). This may imply that moderate exercise not only prevents inflammation-related thrombosis (e.g. atherosclerosis) by reducing platelet activation (sCD40L and P-selectin), but precludes cardiomyopathy or cardiac remodeling by diminishing pro- inflammatory cytokines and matrix metalloproteinase generation.

Inflammation versus thrombogenesis: adiponectin

Circulating concentrations of adiponectin are reduced in the presence of metabolic and cardiovascular disease such as obesity and type 2 diabetes. Adiponectin plays an important role in the control of metabolic dysfunction (e.g. improving glucose homeostasis, insulin sensitivity, and fatty acid oxidation), NO production, and inflammation reduction (via inhibition of NF-κB signaling and CRP/IL-6 generation) (118). Indeed, high leptin and low adiponectin levels lead to dysfibrinolysis. Therefore adiponectin may be bridging inflammatory responses and thrombogenesis as well. Eriksson et al. observed that heavy exercise may suppress leptin concentration, and elevate adiponectin level and fibrinolysis (through reducing PAI-1:Ac); at 6 weeks after exercise, the changes returned to baseline. Hence, exercise can modulate fibrinolytic activity by modifying adiponectin (119). Thus, moderate-to-high intensity aerobic exercise can effectively increase anti-inflammatory and anti-atherogenic properties by elevating adiponectin levels. However, exercise (neither acute nor chronic intensive resistance exercise) without linking to low-grade systemic inflammation changes had no influence on circulating adiponectin levels (120).

Inflammation versus thrombogenesis: platelet and leukocyte interactions

Exercise may cause multicellular activation and thus promote thrombosis. Vigorous exercise-induced pro-inflammatory cytokine (IL-6, CRP) production may lead to augmented oxidative stress (121), enhanced platelet/leukocyte activation (122–124), and greater platelet–leukocyte hetero-aggregation, the trilogy of atherothrombosis. Connolly et al. also noticed that, in a healthy young group, strenuous exercise not only increased inflammation and proliferation, but enhanced anti-inflammation counterbalance (123,125). However, moderate exercise or long-term physical training can diminish pro-inflammatory cytokine release, thereafter ameliorating platelet and/or leukocyte aggregation and adhesion, and reducing plasma fibrinogen levels. The protective mechanism against inflammation and thrombosis is mediated by nitric oxide (NO) (125,126). Meanwhile, resistance exercise training could promote fibrinolytic activity without increasing pro-inflammatory cytokines or procoagulant mediators (83,125).

Additionally, strenuous exercise-induced leukocytosis and leukocyte activation (127) promote fibrinogen binding to platelet glycoprotein IIb/IIIa receptors, facilitate platelet–leukocyte hetero-aggregation, enhance coagulation factors expression or release (e.g. tissue factor on monocyte or factor Va from platelet), and contribute to vascular wall aging (128,129). Elevated catecholamines and leukocyte proteinase (e.g. elastase) following maximal exercise decrease NO availability and antithrombotic protection, increase oxidative stress and tissue damage, finally leading to atherogenesis and/or atherothrombosis in patients with stable angina (130).

Taken together, heavy exercise-mediated oxidative stress, which is mainly due to mitochondrial free radical over-production in skeletal muscle and myocardium, may be atherogenic or atherothrombotic by enhancing lipid peroxidation and inducing endothelial dysfunction. Increased levels of reactive oxygen species (ROS) not only enhance leukocyte adhesion molecule expression and pro-inflammatory cytokine release, but reduce endothelium antioxidant (e.g. NO) and fibrinolysis (e.g. reduced tPA and elevated PAI-1) effects. However, long-term moderate physical training can mitigate vigorous exercise-induced pro-inflammatory and/or prothrombotic effects.

Inflammation versus thrombogenesis: erythrocyte rheology (aggregability and stiffness)

Clinical investigations have demonstrated that enhanced erythrocyte aggregation and blood viscosity correspond to low-grade inflammation in patients with CVD and were frequently accompanied by increases in vascular resistance and thromboembolism (131,132). Increased oxidative stress progressively facilitates erythrocyte neocytolysis/senescence and enhances erythrocyte aggregability/stiffness hances neocytolybitor-1; because of the lack of nuclei and organelles for de novo synthesis of antioxidants. The impaired rheological properties of erythrocytes may further increase vascular resistance and augment subsequent thrombotic risks (133).

In addition, body composition, blood lipid pattern, and plasma viscosity (e.g. fibrinogen levels) also have influence on erythrocyte metabolic and rheological properties. The aggregability and rigidity of erythrocytes may be modified by exercise-dependent blood lipid variation, viscosity alteration, and lactate adaptation (134). Some impairment of blood rheology may be involved in the cardiovascular risk of strenuous exercise, together with changes in hemocoagulatory parameters (135). The viscosities of whole blood and plasma increase in response to a variety of exercise protocols, whereas the deformability and aggregability of red blood cells remain unaltered (i.e. a light exercise (136,137) possesses similar rheological changes to a strong workload) (138,139). Thus, it is likely that changes in hematocrit, red cell rigidity, and plasma viscosity are physiological adaptive modifications during various exercise intensities and do not imply thrombotic risks.

The risk of exhausting exercise workload is more related to other factors, including muscular damage or white cell activation (134). Moreover, repeated and prolonged high shear stresses induced by intensive exercise may disrupt erythrocyte skeleton and disorganize the plasma membrane and, hence, explain the loss of endothelial NO synthase (eNOS) and vascular stiffness after the high-intensity period of exercise (140,141).

However, long-term regular physical training can improve blood fluidity along with hormonal and metabolic alterations, therefore benefit hemostatic balance (142,143). On the one hand, the blood dilutional effect of endurance training may be advantageous in delivering oxygen to the exercising muscles by reducing blood flow resistance, and offsetting dehydration (84). On the other hand, regular training increases the resistance against oxidative stress by attenuating antioxidant depletion in muscle, liver, and erythrocytes during exercise, which can, in turn, preserve the red cell deformability and improve functional capacity by enhancing vasomotor control (144).

In summary, exercise-induced inflammatory alterations may influence coagulation, fibrinolysis, platelet–leukocyte reactivity, and red cell rheology through microparticle production, platelet sCD40L secretion, adiponectin modulation, and nitric oxide formation (Figure 2). Large-scale epidemiological studies have demonstrated that increasing physical activity reduces cardiovascular disease risk. An acute bout of strenuous exercise (e.g. exercise stress test, marathon or > 80% HRmax) enhances oxidative stress (i.e. increases pro-inflammatory cytokines and reduces antioxidant levels), lipid peroxidation, platelet–leukocyte activation, endothelium dysfunction, and erythrocyte aggregation, finally leading to atherothrombotic risks. Additionally, the responses to vigorous exercise are augmented in patients with established cardiovascular disease, or in positive proportion to the aging process.

Figure 2. The training effects of regular (moderate) exercise and low-intensity resistance exercise on inflammatory processes and thrombosis reduction. ↑ = increased; ↓ = decreased; CRP = C-reactive protein; EC = endothelium; IL = interleukin; IL-1ra = interleukin-1 receptor antagonist; NO = nitric oxide; PAI = plasminogen activator inhibitor; PLT = platelet; RBC = red blood cell; sCD40L = soluble CD40 ligand; TF = tissue factor; tPA = tissue plasminogen activator.

However, long-term moderate physical training (60%–79% HRmax, 3–7 times/week, > 8 weeks) ameliorates inflammatory process (through inhibiting NF-κB signaling), improves endothelium function (54–57) and erythrocyte deformability (via increasing NO availability) (142,143), potentiates fibrinolytic advantages (mainly reducing PAI-1 and increasing tPA activities), and eventually would be beneficial to thrombotic risk reduction or prevention. Interestingly, low-volume resistance exercise (low-load (30% of 1 repetitive maximum, i.e. 30% 1RM) or high resistance (80% 1RM) with low repetition) recently has been suggested as part of comprehensive cardiac rehabilitation programs considering the profitable effects on muscle development, glucose/lipid metabolism, and fibrinolytic advantage (82). Therefore, low-load resistance training is similar to endurance exercise, which may reduce inflammation and enhance fibrinolysis, and is a safe exercise program for patients with cardiovascular disease.

Future directions

Physical inactivity has become a global epidemic risk factor for chronic disorders, including cardiovascular disease (CVD). However, too much attention has been placed on the beneficial effects of regular exercise. Adverse responses (i.e. exercise-induced change worsen a risk factor) to regular exercise in cardiovascular and diabetes risk factors occur (145). Hence, while discussing exercise-induced changes in CVD patients, considerable interindividual variability in response to exercise (training) and a more personal tailored exercise (training) program should be borne in mind. Besides, the diverse results and fragmentary information has left many missing links between exercise effects on inflammation and thrombogenesis. Currently evidence on the implications of exercise-induced inflammation for thrombogenesis is relatively weak due to lack of studies assessing causality and the few large-scale case-control interventional studies. Indeed, robust studies directly linking exercise to inflammation and thrombogenesis are still limited. Moreover, various exercise protocols (e.g. resistance or interval exercise) require further exploration. Therefore, standardized exercise programs, larger sample size case-control design, and longitudinal observation are essential.

Declaration of interest: The authors report no conflicts of interest.