Does exercise increase the risk of upper respiratory tract infections?

Transcription

1 Published Online March 31, 2009 Does exercise increase the risk of upper respiratory tract infections? André Moreira,,*,Luís Delgado, Pedro Moreira } and Tari Haahtela Department of Immunology, Faculty of Medicine, University of Porto, Porto, Portugal Immunoallergology, Hospital of São João, Porto, Portugal } Faculty of Nutrition and Food Sciences, University of Porto, Porto, Portugal, and Skin and Allergy Hospital, Helsinki University Central Hospital, Helsinki, Finland Introduction or background: Heavy exercise induces marked immunodepression, which is multifactorial in origin. Evidence showing clinical significance of this immunodepression is scarce. Sources of data: We assessed in a systematic manner whether physical activity or intensity of exercise increase susceptibility to upper respiratory tract infections (URTI). A literature search was performed using the keywords upper respiratory tract infections, athletes, exercise and physical activity. We considered all studies reporting of the effect of exercise, physical activity, sport and training on susceptibility to URTI. A total of 162 publications were identified and 30 studies were eligible (4 descriptive, 18 observational and 8 interventional). The 30 studies included 8595 athletes (5471 runners, 2803 swimmers) and 1798 non-athletes. Areas of agreement: Moderate activity may enhance immune function, whereas prolonged, high-intensity exercise temporarily impairs the immune competence. Athletes, when compared with lesser active individuals, experience higher rate of URTI after training and competitions. In non-athletes, increasing physical activity is associated with a decreased risk of URTI. Areas of controversy: The relationship between exercise and URTI is affected by poorly known individual determinants such as genetic factors, fitness, nutritional status or atopy. Elite athletes may have a decreased susceptibility to URTI. Accepted: March 07, 2009 *Correspondence to: A. Moreira, Department of Immunology, Faculty of Medicine, University of Porto, Al. Prof. Hernâni Monteiro, 4202 Porto, Portugal. andremoreira@ med.up.pt/ amasm@amed.up.pt Growing points and areas timely for developing research: The dose response relationship between immunodepression and risk for URTI during the weeks following heavy exercise. What are the clinically relevant methods to assess exercise-induced immunodepression? Is down-regulation of immunity after intense exercise a protective response to limit inflammation? Is there a role for nutritional or pharmaceutical interventions to reduce risk of URTI? Keywords: athletes/exercise/immunosuppression/respiratory tract infections/ sports British Medical Bulletin 2009; 90: DOI: /bmb/ldp010 & The Author Published by Oxford University Press. All rights reserved. For permissions, please journals.permissions@oxfordjournals.org

2 A. Moreira et al. Background In contrast to moderate or intermittent physical activity, prolonged and intensive exertion causes numerous changes in immunity that possibly reflects physiological stress and suppression. Athletes undertaking regular strenuous exercise may be at increased risk of upper respiratory tract infection (URTI) during periods of heavy exercise and for a couple of weeks following race events. 1,2 It has been suggested that the relationship between exercise and URTIs follows a J-curve, with moderate and regular exercise improving the ability to resist infections 3 5 and heavy acute or chronic exercise decreasing it. 6 8 Exercise-induced immunodeppression has a multifactorial origin, depending on mechanisms related to neuro immune endocrine systems. The evidence shows that prolonged periods of intense training may lead to high numbers of neutrophil and low numbers of lymphocytes in blood counts, 9 11 impaired phagocytosis 12 and neutrophilic function, 13 decreased oxidative burst activity, 12 natural killer cell cytolytic activity (NKCA) 14,15 and mucosal immunoglobulin levels. 16 Intense training may also enhance release of pro-inflammatory cytokines such as tumour necrosis factor alfa, interleukin-1b (IL-1b) and IL-6, followed closely by anti-inflammatory cytokines, such as IL-10 and interleukin-1 receptor antagonist. 17 However, measurements of neutrophil or lymphocyte numbers and NKCA in the blood may not reflect the immunocompetence as a whole. The immune system has many overlapping components, sometimes with shared redundant functions. Although the change of one or more components may affect the host ability to resist infections, it is not possible to predict the overall effects of small to moderate changes in immune parameters on host resistance. The most useful outcome, from a clinical point of view, would be the incidence of URTIs. Studies using multiple doses of exercise are scarce, and the J curve hypothesis has been built based on a composite of results drawn from observational and case series surveys. A recent study, in a large cohort of marathon runners, found no relation between training volume in the 6 months before the race and pre- or post-race infections episodes. 18 Furthermore, increased incidence of URTIs after the marathon was only observed in runners who experienced symptoms in the 3 weeks prior to the race. 18 This suggests that the increased rate of infections among athletes can be caused by strenuous exercise too soon after an infection. The available studies together reinforce the complexity of the causal association between exercise and susceptibility to infections. In this review, we aimed to assess, in a systematic manner, whether physical activity or strenuous exercise increase susceptibility to URTIs. 112 British Medical Bulletin 2009;90

3 Exercise-induced immunodepression Table 1 Characteristics of studies examining the influence of exercise or training and the risk of upper respiratory infections. Study type and reference Design and athletes Exercise or event Outcomes and results Descriptive Green et al. 53 Retrospective; 20 marathon runners aged years Schouten et al. 54 Shephard et al. 55,56 Malm 35 Retrospective; 199 adults aged years Retrospective; 750 older age-classed masters competitors, years 16-year follow-up of one elite distance runner Observational Peters and Bateman (1983) 31 Case control; 150 marathon runners and age matched controls Seyfried et al. 31 Prospective; 2743 swimmers and 1794 non-swimmers aged 0 70 years Linde (1987) 30 Cross-sectional; 42 elite orienteer s aged years and 41 matched controls Strauss et al. 57 Prospective; 87 university wrestlers, swimmers and gymnasts None Self reported physical activity during previous 3 months The weekly time devoted to training, competition and exercise-related travel was h Daily training log for 16 years 1982 two Oceans Marathon Nine out of 20 felt that marathon running increased their self-perceived resistance to URTIs; 1 had opposite feeling For women a negative very week correlation association between the reported incidence of URTI symptoms and the level of sports activity in the previous 3 months was observed Increased risk of self-reported URTI if distance above km/week Inverse association between exercise load and risk of infections Significantly more self-reported symptoms of URTI 1 week after the race in runners compared with controls Incidence of self-reported URT symptoms higher in fastest runners, whit no differences between slow runners and controls 3 months usual training Increased self-reported incidence of URTIs in swimmers Median of 8 h week of training 8 weeks of their competitive season Orienteer s reported more URTIs episodes than controls No significant differences in self-reported incidence of URTI symptoms between groups were observed Continued British Medical Bulletin 2009;90 113

4 A. Moreira et al. Table 1 Continued Study type and reference Design and athletes Exercise or event Outcomes and results Nieman et al. 22 Prospective; 273 marathon runners Nieman et al. 8 Prospective; 2311 marathon runners Heath et al. 24 Prospective 12 months; 530 runners Verde et al. 25 Prospective; 10 highly trained distance runners, years Linenger et al. 58 Prospective; 482 trainee-months; subjects aged years Gleeson et al. 28 Prospective 7-month follow-up; 26 elite swimmers aged years and 12 controls Mackinnon and Hooper 59 Prospective 4-week follow-up; 24 elite swimmers aged years 2-month training period prior to a 5 km, 10 km, or half-marathon race LAM Usual daily running habits kept 38% increment of training over a 3-week period US Navy SEAL training h of training per week for 7 months 4 week of intensified training Self-reported URTI symptoms were more common in runners training more than 15 miles per week No increase in infectious episodes during the week following the run compared with the week prior to the race Training greater than or equal to 97 versus less than 32 km/week, doubled the risk of self-reported URTI during the 2 month prior to the marathon Runners who attend the marathon reported six times more infection episodes on the week following the run than those who did not participate The running dosage was a significant risk factor for self-reported URTI symptoms 2 out of 10 developed rhinovirus infection Incidence of URTI symptoms was 14.7 cases per 100 trainee-months No significant differences in infection rate between athletes and controls clinically assessed 10 of 24 swimmers exhibited URTI during the study No difference in glutamine levels between athletes who developed URTI and those who did not Continued 114 British Medical Bulletin 2009;90

5 Exercise-induced immunodepression Table 1 Continued Study type and reference Design and athletes Exercise or event Outcomes and results Weidner et al. 60 Case control; 50 moderate fitness adults, years were inoculated with rhinovirus after allocation on exercise (n ¼ 34) or sedentary (n ¼ 16) group Gleeson et al. 29 Prospective 3-month follow-up; 22 elite swimmers aged years Gleeson et al. 21 Prospective 30-day intensive training; 14 elite swimmers aged years Matthews et al. 61 Fricker et al. 34 Ekblom et al. 18 Prospective 12 month of follow-up; 547 healthy adults (49% women) aged years Prospective; 20 highly trained international levels male distance runners Prospective 3 week before and 3 week after the run; 1694 runners Moderate exercise training: 40 min every other day at 70% of HR reserve for a 10-day period h of training per week for 3 months m swam per day, for 30 days Five evaluations with three 24-h physical activity recalls per evaluation A 4-month winter training period in the southern hemisphere Stockholm Marathon 2000 No effect on severity or duration of rhinovirus caused URTI No relationship between clinically assessed URTIs and systemic markers of immunodepression Significant relationship between EBV serology status and upper respiratory tract symptoms Moderate levels of physical activity were associated with a reduced risk for URTI There were no substantial relationships between mean weekly training mileage, intensity or training load and the number of illnesses reported; neither submaximal nor maximal running performance was significantly affected by the presence of illness Younger age and pre-race health status were factors explaining post-race reported URTIs incidence No relation between training volume 6 months before the race, finishing time and socioeconomic and demographic factors and pre- or post-race reported URTIs Continued British Medical Bulletin 2009;90 115

6 A. Moreira et al. Table 1 Continued Study type and reference Design and athletes Exercise or event Outcomes and results Spence et al. 27 Prospective 5-month; 32 elite, 31 competitive triathletes and cyclists, and 20 sedentary controls; aged years Intervention Nieman et al. 26 Nieman et al. 4 Peters et al. 23 Peters et al. 32 Chubak et al. 62 Kekkonen et al. 38 Randomized, controlled; 50 mildly obese, sedentary women, years Randomized, controlled; 32 sedentary, elderly women, years Randomized, controlled study; 84 runners and 73 controls Randomized, controlled study; 178 runners and 162 sedentary controls Randomized, controlled study; 115 overweight and obese, sedentary, postmenopausal women Randomized double-blind controlled trial; 141marathon runners Usual training and competition 15-week exercise training, five 45-min sessions/week, walking at 60% HR 12-week exercise training, five 40-min sessions/week, walking at 60% HR or calisthenic mild range-of-motion and flexibility movements 90-km Comrades Ultramarathon 90-km Comrades Ultramarathon 45 min of moderate-intensity exercise 5 days per week for 12 months versus once-weekly, 45-min stretching sessions 3-month training period before and 2 weeks after the Helsinki Marathon 2003 Higher rate of URI among elite athletes than competitive athletes Pathogens were isolated in fewer than 30% of URI cases Exercise halved the days with incident URTI symptoms Improvement in cardiorespiratory fitness was correlated with a reduction in reported URT symptoms Lower incidence of symptoms of URTIs confirmed by investigator in the walkers (50%) compared with calisthenic (21%) Incidence of self-reported URTI symptoms higher in runners who trained hardest Significantly more symptoms of self-reported URTI after the race in the placebo group than in the sedentary controls Risk of URTIs overall did not differ; however, over the risk of colds decreased in exercisers relative to stretchers LGG had no effect on the incidence of respiratory infections or GI-symptom episodes during the training period and the 2 week after the marathon, but shortened the duration of GI-symptom episodes Continued 116 British Medical Bulletin 2009;90

7 Exercise-induced immunodepression Table 1 Continued Study type and reference Design and athletes Exercise or event Outcomes and results Henson et al. 63 Cox et al. 40 Randomized double-blind controlled trial; 39 ultramarathon runners Randomized double-blind, placebo-controlled, cross-over trial; 20 marathon runners Summary Four descriptive; 18 observational; and 8 intervention studies; 8595 athletes (including 5471 runners and 2803 swimmers) and 1798 non-athletes 160 km Western States endurance run 2-month training period, about 100 km week studies on physical activity and 8 on exercise From calisthenic mild range-of-motion and flexibility movements to ultramarathon running 1000 mg/day quercetin for 3 weeks before, during, and 2 weeks after the race had no effect on illness rates, perturbations in leukocyte subset counts, or decreases in granulocyte respiratory burst activity and salivary IgA Lactobacillus fermentum VRI 003 (PCC) was associated with a substantial reduction in the number of days and severity of respiratory illness In non-athletes, increasing physical activity is associated with a decrease in the risk of URTIs Studies comparing higher and lesser active individuals suggest that athletes experience higher rate of URTI episodes after training and competing than controls; Studies which compared the incidence of URTI symptoms in same subjects during training and competition suggest a dose response relationship between risk of infection and cumulative training; however, a significant increase in susceptibility to infections is incompatible with sustained training at an elite level Younger age and pre-race infection episode seem to increase the risk of URTI after race EBV, Epstein Barr virus; HR, heart rate; RV, Rhinovirus; URT, upper respiratory tract. British Medical Bulletin 2009;90 117

8 A. Moreira et al. Methods The primary method to locate potentially eligible studies was a computerized literature search in the MEDLINE database, from inception to December 2008, using the following search keywords: (upper respiratory tract infections OR respiratory infections) AND (athletes OR exercise OR physical activity) AND (clinical and trial or clinical trials or clinical trial). Descriptive ( prevalence surveys, case series, surveillance data and analysis of registries), observational (cohort, case control, cross-sectional and ecological) and intervention studies (including randomized, quasi-randomized and non-randomized trials) were retrieved. We considered all studies reporting on the effect of physical activity, sport or training on susceptibility to URTIs. Studies including healthy, athletes or non-athletes, performing acute strenuous exercise or engaged in training programs, were considered. No restrictions for sex or age were made for participants. The primary outcome measure was URTIs either self-reported or medically assessed. One hundred and sixty-two publications were identified as potentially relevant. All of the titles and abstracts that were identified by the search and that appeared relevant were selected for a full-text review. Trials were assessed for inclusion by looking at the objective and methods section of each paper while remaining unaware of the study s results, conclusion and authors. The reference lists of all primary studies were reviewed to identify trials not retrieved by the electronic and manual searches. Assessment of quality of observational studies was performed using the Newcastle Ottawa scale star system, 19 in which a study is judged on three broad perspectives: the selection of the study groups, the comparability of the groups and the ascertainment of either the exposure or outcome of interest was used for the quality rating of observational studies. Quality rating of intervention trials was assessed with particular emphasis on allocation concealment and randomization. 20 Results We retrieved 30 studies, comprising 4 descriptive, 18 observational and 8 randomized or controlled reports, that included a total of 8595 athletes, including 5471 runners and 2803 swimmers and 1798 nonathletes (Table 1). The quality rating gave a median score of 0.5 out of 4(min¼ 0, max ¼ 1), 5 out of 9 (min ¼ 1, max ¼ 8) and 9 out of 15 (min ¼ 3, max ¼ 12), respectively, for descriptive, observational and intervention trials (Table 2). Studies reported on different patterns of physical activity, ranging from mild training programs for elderly people to preparation for 118 British Medical Bulletin 2009;90

9 Exercise-induced immunodepression international competition events by elite athletes. Studies with athletes rarely reported baseline fitness, or used objective measures of activity, such as frequency, intensity and duration, and the level of compliance to the training. Studies referring to the effects of acute bouts of exercise provided none or little information on the intensity of effort. Evidence of upper respiratory infection based on objective criteria, such as the detection of the virus in isolates or seroconversion, was only performed in one trial. 21 In others, information was obtained from the retrospectively analysis of medical records or the individual s self-reported symptoms or incidence of respiratory symptoms assessed by questionnaire. No severity assessments were performed. None reported on the impact of confounding conditions on symptoms, such as allergies and asthma, or whether increased diagnoses of infection during periods of heavy training or competition resulted from higher medical surveillance. Adjustment for other factors such as the likelihood of contacts with infection, type of sports and training environment, overall levels of stress, general health, nutritional status and the use of dietary supplements that influence the risk of upper respiratory infection was rarely done. The few studies which compared the incidence of URTI symptoms in same subjects during training and competition suggest a dose response relationship between risk of infection and cumulative training affected 4,18,22 27 by age, fitness and health status. In the study of Nieman et al. 22 URTI symptoms were more common in runners who trained more, although no increase in infectious episodes in the week after the run was seen for these athletes. These observations were supported in three other trials where the incidence of URTI symptoms was higher in runners who trained hard In the study by Heath et al., 24 the odds for an URTI doubled or tripled for running mileage between 486 and 865 or above 1388 miles, respectively. These findings were challenged by the 2000 Stockholm City Marathon study, where no relationship between training volume for the 6 months prior to the race, finishing time, socioeconomic and demographic factors and pre- or post-race reported URTIs was observed. 18 Age and pre-race health status were the determinants of post-race URTIs incidence being the younger runners or athletes with pre-race infection episode the ones more prone to experience URT after the race. 18 Subjects fitness may also affect the risk of URTI. While, in high fit subjects, such as international level competitors, the risk of URTI seems to equals the controls, 21,28,29 in mid-aged or elderly sedentary women increases in the level of physical activity halves the days with incident URTI symptoms. 4,26 Case control studies comparing higher and lesser active individuals 1,27,30 32 suggest the former are at an increased risk of URTI incidence. However, these studies are heterogeneous, which included runners, British Medical Bulletin 2009;90 119

10 A. Moreira et al. Table 2 Quality rating of included studies. Study type and reference Selection Comparability Outcome Descriptive* Green et al. 53 þ Schouten et al. 54 Shephard et al. 55,56 þ Malm et al. 35 Observational Peters et al. 1 þ þ Seyfried et al. 31 þþ þ Linde 30 þ Strauss et al. 57 þ Nieman et al. 22 þþ þ Nieman et al. 8 þþ þ Heath et al. 24 þþ þ Verde et al. 25 þþ þ þþ Linenger et al. 58 þþ þ þþ Gleeson et al. 28 þþþ þ þþ Mackinnon and Hooper 59 þþþ þ þþ Weidner et al. 60 þþ þþ þþþ Gleeson et al. 29 þþ þ þþþ Gleeson et al. 21 þþ þ þþþ Matthews et al. 61 þþ þ þþþ Fricker et al. 34 þþþ þ þþ Ekblom et al. 18 þþþ þ þþ Spence et al. 27 þþþþ þ þþ Intervention Nieman et al. 26 þþ þ þþþ Nieman et al. 4 þþþ Peters et al. 23 þþ þ þþþþþþ Peters et al. 32 þþ þ þþþþþþ Chubak et al. 62 þþ þ þþþþþþ Kekkonen et al. 38 þþ þ þþþþþþþþþ Henson et al. 63 þþ þ þþþþþþþþþ Cox et al. 41 þþ þ þþþþþþþþþ *Descriptive studies (a study can be awarded a maximum of one þ for each numbered item within the selection and exposure categories, a maximum of two þ can be given for comparability). 1 ¼ Selection: case definition: adequate definition of the case with independent validation (þ), cannot tell (0); representativeness of the cases: obviously representative series of cases (þ), potential for selection biases or not stated (0); 2 ¼ Comparability: comparability of cases and controls on the basis of the design or analysis: study controls for subjects immune systems stressors, e.g. neuro-humoral, individual fitness or nutritional status (þ), no description (0); study controls for any extra subject additional risk factors, e.g. travelling to and from sports venues, large scale participation events, or seasonal variations in infection diseases (þ), no description (0). 3 ¼ Reporting: ascertainment of exposure: secure record (þ); written self-report (0); non-response rate: same rate for both groups (þ), rate different and no designation (0). Observational studies including case control and chort studies: case control studies (a study can be awarded a maximum of one þ for each numbered item within the Selection and Exposure categories, a maximum of two þ can be given for Comparability). 1 ¼ Selection: case definition: adequate definition of the case with independent validation (þ), record linkage or based on self-reports (0) cannot tell (0); representativeness of the cases: consecutive or obviously representative series of cases (þ), potential for selection biases or not stated (0); selection of controls: community controls (þ), no description (0); definition of controls: no history of disease (þ), no description of source (0). 2 ¼ Comparability: comparability of cases and controls on the basis of the design or analysis: study controls for subjects immune systems stressors, e.g. neuro-humoral, individual fitness or nutritional Continued 120 British Medical Bulletin 2009;90

11 Exercise-induced immunodepression Table 2 Continued status (þ), no description (0); study controls for any extra subject additional risk factors, e.g. travelling to and from sports venues, large scale participation events, or seasonal variations in infection diseases (þ), no description (0). 3¼ Exposure: ascertainment of exposure: secure record or structured interview where blind to case/control status (þ); interview not blinded to case/control status or written self-report (0); same method of ascertainment for cases and controls: yes (þ), no (0); non-response rate: same rate for both groups (þ), rate different and no designation (0). Cohort studies (a study can be awarded a maximum of one þ for each numbered item within the Selection and Outcome categories, a maximum of two þ can be given for Comparability). 1 ¼ Selection: representativeness of the exposed cohort: yes, representative of the average (þ), no (0); selection of the non exposed cohort: drawn from the same community as the exposed cohort (þ), no description (0); ascertainment of exposure: secure record or structured interview (þ), no description (0); demonstration that outcome of interest was not present at start of study: yes (þ), no (0). 2 ¼ Comparability: comparability of cases and controls on the basis of the design or analysis: study controls for subjects immune systems stressors, e.g. neuro-humoral, individual fitness or nutritional status (þ), no description (0); study controls for any extra subject additional risk factors, e.g. travelling to and from sports venues, large scale participation events, or seasonal variations in infection diseases (þ), no description (0). 3 ¼ Outcome: assessment of outcome: independent blind assessment or record linkage (þ), self-report or no description; follow-up long enough for outcomes to occur: yes (þ), no (0); adequacy of follow up of cohorts: complete follow up or subjects lost to follow up unlikely to introduce bias (þ), no description of those lost (0) Intervention studies (a study can be awarded a maximum of one þ or þþ for each numbered item within the Selection or Procedures categories, a maximum of two þþ can be given for Comparability): 1 ¼ Selection: all eligible subjects with number of and reason for exclusions given: yes (þ), no (0); homogeneity of type-of-sports practiced: yes (þ), no (0); similar distribution between groups at baseline reported (þ), not reported/not applicable (0); 2 ¼ Comparability: comparability of cases and controls on the basis of the design or analysis: study controls for subjects immune systems stressors, e.g. neuro-humoral, individual fitness or nutritional status (þ), no description (0); study controls for any extra subject additional risk factors, e.g. travelling to and from sports venues, large scale participation events, or seasonal variations in infection diseases (þ), no description (0). 3 ¼ Procedures: randomisation: concealed randomisation (computer, centralised, etc) (þþ); potentially manipulable (sealed envelope, date of admission, medical records, birth date, etc) (þ); can t tell (0); blinding: double blind (research team/athletes) (þþ); single blinded (þ); unblinded/can t tell (0); intention-to treat analysis: reported (þ), not reported (0); allocation concealment: adequate (þþ), unclear or inadequate (þ),not used (0); Intervention: described explicitly (þ), not reported (0); assessment of outcome: independent blind assessment or record linkage (þ), self-report or no description; follow-up long enough for outcomes to occur: yes (þ), no (0); adequacy of follow up of cohorts: complete follow up or subjects lost to follow up unlikely to introduce bias (þ), no description of those lost (0) triathletes and swimmers, and differences of lifestyle between athletes and controls, such as consumption of nutrional supplements, were not registered. A report of 5-month follow-up showed lower rates of illness in recreationally competitive athletes, and substantially higher rates in elite triathletes and cyclists. 27 In another study, where comparisons were made between marathon runners who attend and those who miss the marathon by reason other than URT disease, running more than 97 km per week doubled the risk of infection during the training period. 8 Also, runners attending the marathon reported more infection episodes on the week following the run than those who did not participate 8 or than sedentary controls. 32 British Medical Bulletin 2009;90 121

12 A. Moreira et al. Discussion Reviewer s conclusions Studies looking at the impact of exercise or physical activity on susceptibility to infection varied widely in respect to subjects, exercise load and methods. Pooling data and performing a meta-analysis under such a high degree of heterogeneity are not advisable. Although the J -shaped curve hypothesis relating the amount of exercise and risk of disease has been accepted by athletes, coaches and scientists, the available evidence is insufficient to support it. However, there appears to be an association between increased susceptibility to infection and strenuous effort, such as marathon running. Subject s individual s factors, among which age and pre-event health status seems the most important, certainly play a role. We hypothesize that in high fit subjects, such as international level competitors, the association between exercise load and risk of URTIs may tend to flatten (Fig. 1). Exercise and URTIs Many studies are now available to explain how acute and exhaustive exercise may increase the susceptibility to URTIs (for a review see Gleeson 33 ). Intense exercise is known to decrease the expression of Fig. 1 A model of the relationship between upper respiratory tract infection (URTI) risk and intensity of exercise. The evidence suggests a J-shaped curve in the relationship between risk of URTI and exercise intensity in lower fitted individuals. As fitness increases, odds favouring the infection would decrease in the extremes of the curve which tends to a flat line. 122 British Medical Bulletin 2009;90

13 Exercise-induced immunodepression toll-like receptors and to increase cortisol, epinephrine and IL-6 production leading to an impaired cell-mediated immunity and low-grade inflammation, via a decreased macrophage and Th-1 cell cytokine production. 33 Epidemiologic reports suggest that athletes engaging in strenuous exercise, such as a marathon run or heavy training, are at increased risk of URTI. In a large study, including more than 2000 marathon runners who varied widely in running ability and training habits, the odds for a self-reported URTI was 6-fold higher in marathoners who participated when compared with those who trained but missed the race. 8 Subsequent studies from Peters et al. 23,32 have partially supported this observation. In a double-blind, placebo, controlled trial, assessing the effects of antioxidants supplementation, symptoms of URTI after the race, is more common in the placebo group than in the sedentary controls, 32 and runners who train harder report higher incidence of URTI symptoms. 23 However, the direct dose response relationship between exercise load and risk of URTI is not consistent and some studies suggest either no change, or even a slight reduction, in the risk of sickness after a run. It has been shown no increased prevalence of URTI in marathoner runners during the week after 5, 10 and 21.1-km events when compared with the week before, with athletes training more than 25 km per week being protected from infections, although non-significant. 22 A similar observation was reported by Heath et al. 24 with the odds ratio for URTI increasing with averaged running distance until 1388 miles per year, but decreased thereafter. These apparent discrepant findings may reflect differences in subjects and methods among studies. However, two studies give important data to reconcile these observations. In the Stockholm marathon study, if runners with at least one self-reported pre-race infection episode were excluded from the analysis, the post-race infection episode incidence would be equal to the selfreported incidence before the marathon. 18 Neither pre-race training volume nor running time, nor gender or social status could explain the incidence of post-race infection episode. However, faster finishing time in relation to pre-race training status may be a risk factor, especially in younger runners. 18 Additionally, a longitudinal observational study of long-distance runners with serial monitoring of training loads and clinical patterns of illness found no effect of training mileage, intensity and load on the incidence of respiratory illness. 34 These data suggest a major role for the determinants of the risk of URTIs being dependent on the subject and not on the exercise load, emphasizing the need in the international competitor athlete of not only an extremely prepared physique system but also of an immune system able to withstand infections during severe physiological and psychological stress. 35 British Medical Bulletin 2009;90 123

14 A. Moreira et al. Physical activity and URT The evidence suggesting moderate physical activity may to be beneficial in decreasing URTI risk is limited. Two studies, both randomized, but including a small number of subjects, suggest that women engaging in a 12- or 15-week cardio-respiratory training are more likely to be protected from an URTI than those who do not exercise regularly. 4,26 Physical activity is an important health-related variable, albeit difficult to measure. Additionally, to self-report activities using questionnaires and/or diaries where information on all forms of activity is recorded each day, objective assessments include measures of fitness by direct and indirect maximal oxygen uptake assessments, measures of energy expenditure using direct calorimetry with doubly labelled water and motion sensors that measure activity in one or more planes of movement. Engagement in physical activity decreases the incidence of coronary heart disease, diabetes, hypertension, some cancers and osteoporosis. 36 In addition to its preventive effect, physical activity is recommended in the treatment of several chronic diseases. 36 Nutritional and dietary countermeasures to exercise associated URTIs Four studies, 23,32,37,38 including 358 marathon runners, addressed the efficacy of nutritional supplements decreasing the incidence URTIs. In the study of Himmelstein et al., 37 there were no significant differences in the self-reported incidence of URTIs after exercise when the subjects consumed 1000 mg of vitamin C daily for 2 months prior to and 1 month after the marathon. Two other studies assessed the impact of vitamin C supplementation, either alone or combined with vitamin E or beta-carotene, on the URTI symptoms after the race. 23,32 In the study of Peters et al., 23 the intervention group consumed 600 mg of vitamin C for 21 days prior to the ultra marathon, and the URTIs were assessed for 14 days after the race. In this study, the number of runners of the placebo group reporting development of URTI symptoms after the race was 68%, which was significantly more than that reported by the vitamin C-supplemented group. In the study of Peters et al., 32 athletes were supplemented with vitamin C alone or combined with vitamin E or beta-carotene during the 21 days prior to the race. Significantly, more symptoms of URTI in the placebo group were observed compared with vitamin C or combinations of vitamins C, E and beta-carotene. In fact, vitamin C alone was as effective as when combined with vitamin E or beta-carotene in preventing URTI symptoms. When the data from the studies were pooled, the rate ratio for URTIs after vitamin C supplementation against placebo was halved British Medical Bulletin 2009;90

15 Exercise-induced immunodepression Two other studies evaluated the effect of probiotics consumption in the prevention of infection episodes in athletes. In a randomized, controlled trial, enrolling 139 marathon runners, supplementation with Lactobacillus rhamnosus GG (LGG) for 3 months prior to the Helsinki city marathon, had no effect on the incidence of respiratory infections or gastrointestinal symptom episodes during the training period and the 2 weeks after the marathon. 38 Also, LGG supplementation did not prevent the increase of allergic inflammatory markers during the pollen season, or the eosinopenia induced by the marathon. 39 During a 1-month treatment with the probiotic Lactobacillus fermentum VRI-003, in 20 trained long-distance runners, a significant reduction in the number of days of respiratory illness symptoms, and a trend towards a lower severity of illness was observed. 40 The potentially beneficial effect of the probiotic administration on the incidence of respiratory illness is possibly linked to an enhancement of systemic immunity. LGG appeared to have a slight anti-inflammatory effect reflected through a decrease in IL-6 production, which is the main inducer of C-reactive protein and a decrease in the number of monocytes, which are one of the major producers of IL As different strains of probiotics have different immunostimulating capacities, these findings cannot be generalized to other strains. Consumption of carbohydrate-rich beverages during exercise appears to attenuate some of the immunosuppressive effects of prolonged exercise, but confirmation of any clinical significance in this awaits further research. 20 The plasma concentration of the conditionally essential amino acid glutamine, the most abundant in the human body, is lowered by stress or strenuous physical activity. This is most likely due to increased cellular needs that exceed the supply available from diet or muscle. Plasma glutamine responses are characterized by increased levels during exercise followed by decreased levels post-exercise and during the recovery period. This has led to the glutamine hypothesis as one explanation of immunodepression after exercise (for a review see Castell 42 ). It has been postulated that restoring glutamine levels after prolonged exercise to physiological levels may help the immune system to resist infections. Although oral provision of glutamine, or branched chain amino acids as a glutamine precursor, abolishes the post-exercise plasma glutamine decrease, it has no consistent effect on lymphocyte or neutrophil counts, salivary IgA, oxidative burst activity, NKCA or plasma IL-6 after exercise. 20 The available data do not provide evidence to support the glutamine hypothesis. Taken together, at the present, there is no evidence to support a role for any other nutritional supplements in preventing exercise-induced immune suppression or in protecting against infections. British Medical Bulletin 2009;90 125

16 A. Moreira et al. Allergic athletes and URT An increasing proportion of young athletes is atopic, i.e. show signs of IgE-mediated allergy which is, along with the sport event, a major risk factor for asthma and respiratory symptoms in athletes. 43,44 The relative importance of allergy is on increase, also because pollen exposure may become more prolonged and intense with global warming. 45 A mixed type of eosinophilic and neutrophilic airway inflammation seems to affect especially swimmers, ice-hockey players and crosscountry skiers. The inflammation may represent a multifactorial aggression, in which both allergic and irritant mechanisms play a role. In allergic athletes, high-level competition seems to exacerbate at least some components of the allergic immune response, such as airway hyper responsiveness and airway inflammation. 46 The question remains about how excessive exercise affects the Th1 Th2 balance. If exercise drives a Th2 response then a more difficult to control phenotype in the elite allergic athlete may be expected, however, we found no studies yet addressing this issue. Limitations of the included studies We performed a quality rating of the included trials. The use of such scores can be arguable and potentially misleading, 47 reflecting the difficulties in determining fair quality rates. Assessment of quality of observational studies is more difficult than assessment of quality of randomized trials and other experimental studies. Quality assessment methods for observational studies have not been well worked out and, although several assessment scales and checklists are used, none of them have been fully validated or shown to include criteria that are associated with the effect size in empiric studies. Our conclusions are limited by the low-quality score of included studies, particularly among descriptive and observational studies. The J-curve hypothesis has been based mainly on epidemiological studies where a selection bias may have occurred. 1,8,23,31 In the Los Angeles marathon (LAM) study, only 47% of runners responded, being possible those were the ones who developed symptoms, and so increasing the true incidence of disease. Also, it has been claimed that if comparisons are made between athletes and sedentary individuals or between elite athletes at different points of their training, medical surveillance may be increased during periods of intensive training and at times of competition increasing the likelihood of diagnosis of URTI. 48 Studies had poor comparability of cases and controls. No control for subjects or context dependent factors such as individual fitness or 126 British Medical Bulletin 2009;90

17 Exercise-induced immunodepression nutritional status, or the effects of large scale participation events were performed. Another limitation common to all studies assessing URTIs is the definition of aetiology. It cannot be excluded that a post-exercise bronchial and nasal hyper-reactivity can be reported as an infection episode after intense exercise. 45,49 The exercise-induced immunodepression offer some biological plausibility to the hypothesized J-shaped relationship between physical activity and susceptibility to infection. However, evaluating the suppression of one immunological parameter at a time runs the risk of disregarding changes in host resistance as a whole. This reductionist approach, taken in most of the studies, fails to bring to light the impact of the interventions on overall host resistance. A validated mathematical model should be used to estimate and predict the relationship between multiple immunological parameters and host resistance. 50 In addition, clinical endpoints, such as incidence of properly diagnosed infections, should be used. Very few studies addressed physical activity levels. Physical activity is defined as the behaviours that result in any movement contributing to human total energy expenditure. 51 It includes all large muscle movement, for whatever purposes, carried out throughout the day. Exercise is a subset of total physical activity, being purposive and repetitive movements with the aim of improving measurable cardio-respiratory or other dimension of fitness. Exercise is usually comprised of more structured physical activities, often performed at a vigorous intensity. So, physical activity is a complex multidimensional set of behaviours, with possible measurements made of its duration, frequency, intensity or setting. This has made assessment in trials difficult to perform. Limitations and strengths of the review In common with all systematic reviews, we may have included studies in which the outcomes such as physical activity or exercise and characteristics of the human subjects were too dissimilar for comparison, resulting in questionable conclusions. The subjects of included studies were from different sports or non-fit individuals, exercising in different environmental and stress conditions making any generalizations open to question. Another problem was the lack of randomized, controlled trials assessing the effect of different levels of physical activity in incident infections those studies are extremely, almost impossible, to perform. The present study has an important strength. Because of a comprehensive search strategy, omission of important trials seems unlikely. British Medical Bulletin 2009;90 127

18 A. Moreira et al. Implications for practice and future research Besides improving methodological issues, such as randomisation procedures, allocation concealment, selection of subjects and reporting of adverse events, future studies should make an effort to objectively assess physical activity and exercise intensity. Future investigations should strive to consider individuals with allergic conditions and to include methods of assessing the severity, duration and treatment of upper respiratory infections. These would ideally be based on objective criteria, such as the detection of the virus in isolates or seroconversion. Trials should also refer to the exercise training strategy as exercise itself is an important immunomodulator. 52 Questions to be addressed include: (i) Are the athletes who show more immunodepression more prone to URTIs during the weeks following exercise? (ii) Which are the relevant outcomes to assess exercise-induced immunodepression? (iii) Is down-regulation of nonspecific immunity after intense exercise a normal protective response, with mild immunodepression being an attempt to limit inflammation? (iv) When should this be considered pathological? (v) What are the differences between healthy and illness prone athletes? (vi) What is the efficacy of nutrional or pharmaceutical interventions as countermeasures to incident URT? Conclusion In conclusion, numerous studies have shown an inverse relationship between exercise workloads and function of the immune system. Does exercise increase the risk of URTIs? Moderate activity may enhance immune function above sedentary levels, whereas excessive amounts of prolonged, high-intensity exercise impair the immune competence. We hypothesize this relationship is most likely affected by individual determinants such as fitness, nutritional status or atopy. It is possible that among elite athletes, the relationship between exercise load and immune dysfunction tends to flat, while the J-curve relationship would be seen in the less fit subjects. However, evidence supporting clinical translation of this immunodepression to confirmed illness is lacking and should be subject of research in future studies. Acknowledgment We thank Miguel Pimenta for assistance drawing the figure. Conflict of interest: All authors contributed to the conception, design of the study and interpretation of results. A.M. contributed to data 128 British Medical Bulletin 2009;90

19 Exercise-induced immunodepression acquisition and drafted the article. All authors approved the final version for publication. References 1 Peters EM, Bateman ED (1983) Ultramarathon running and upper respiratory tract infections. An epidemiological survey. S Afr Med J, 64, Peters EM (1990) The effect of altitude on the incidence of upper respiratory tract infections following participation in ultramarathon events (Abstract). S Afr Sports Med, 5, Nieman DC, Davis JM, Brown VA et al. (2004) Influence of carbohydrate ingestion on immune changes after 2 h of intensive resistance training. J Appl Physiol, 96, Nieman DC, Henson DA, Gusewitch G et al. (1993) Physical activity and immune function in elderly women. Med Sci Sports Exerc, 25, Nieman DC, Nehlsen-Cannarella SL, Henson DA et al. (1998) Immune response to exercise training and/or energy restriction in obese women. Med Sci Sports Exerc, 30, Nieman DC (1994) Exercise, upper respiratory tract infection, and the immune system. Med Sci Sports Exerc, 26, Castell LM, Poortmans JR, Newsholme EA (1996) Does glutamine have a role in reducing infections in athletes? Eur J Appl Physiol Occup Physiol, 73, Nieman DC, Johanssen LM, Lee JW, Arabatzis K (1990) Infectious episodes in runners before and after the Los Angeles Marathon. J Sports Med Phys Fitness, 30, Nieman DC, Nehlsen-Cannarella SL, Fagoaga OR et al. (1998) Effects of mode and carbohydrate on the granulocyte and monocyte response to intensive, prolonged exercise. J Appl Physiol, 84, Nieman DC, Simandle S, Henson DA et al. (1995) Lymphocyte proliferative response to 2.5 hours of running. Int J Sports Med, 16, Pedersen BK, Bruunsgaard H, Klokker M et al. (1997) Exercise-induced immunomodulation possible roles of neuroendocrine and metabolic factors. Int J Sports Med, 18 (Suppl 1), S2 S7. 12 Nieman DC, Fagoaga OR, Butterworth DE et al. (1997) Carbohydrate supplementation affects blood granulocyte and monocyte trafficking but not function after 2.5 h or running. Am J Clin Nutr, 66, Nieman DC, Nehlsen-Cannarella SL, Fagoaga OR et al. (2000) Immune function in female elite rowers and non-athletes. Br J Sports Med, 34, Nieman DC (1997) Immune response to heavy exertion. J Appl Physiol, 82, Shinkai S, Kurokawa Y, Hino S et al. (1993) Triathlon competition induced a transient immunosuppressive change in the peripheral blood of athletes. J Sports Med Phys Fitness, 33, Nieman DC, Henson DA, Fagoaga OR et al. (2002) Change in salivary IgA following a competitive marathon race. Int J Sports Med, 23, Nieman DC, Henson DA, Smith LL et al. (2001) Cytokine changes after a marathon race. J Appl Physiol, 91, Ekblom B, Ekblom O, Malm C (2006) Infectious episodes before and after a marathon race. Scand J Med Sci Sports, 16, Wells G, Shea B, O Connell D et al. (2008) The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. University of Ottawa Available from (cited 2008 December 31). 20 Moreira A, Kekkonen RA, Delgado L, Fonseca J, Korpela R, Haahtela T (2007) Nutritional modulation of exercise-induced immunodepression in athletes: a systematic review and meta-analysis. Eur J Clin Nutr, 61, Gleeson M, Pyne DB, Austin JP et al. (2002) Epstein Barr virus reactivation and upperrespiratory illness in elite swimmers. Med Sci Sports Exerc, 34, British Medical Bulletin 2009;90 129