The Effects of Acute and Chronic Exercise on ...

REVIEW. A.RTICLE

Sports Mcdicinc 1 l (3): 18.1-201. 1991 0 l 11-l h41/9 1 /0003-0 I X3/$09.50/0 0 Adl\ Inlcl.nalional Lirnitcd. All 1.1ghts rc5r1-\rd S P 0 1 003A

The Effects of Acute and Chronic Exercise on Immunoglobulins David C. Nieman and Sandra L. Nehlsen-Cannarella Department of Health, Leisure and Exercise Science, Appalachian State University. Boone. North Carolina. and Immunology Center, Departments of Surgery, Pathology, and Microbiology. Loma Linda University Mcdical Ccntcr. Loma Linda, California. USA

Contents Summary I. Structure and Function of Immunoglobulins I . I B Cell Activation 2. Measurement of Serum lrnmunoglobulins 3. Effects of Exercise on lmrnunoglobulins 3.1 Effects of Acute Exercise on lmmunoglobulins 3.1.1 Acute Graded-Maximal Exercise 3.1.2 Acute Intense Submaximal Exercise 3.1.3 Acute Moderate Submaximal Exercise 3.2 Effects of Exercise Training on Serum lmrnunoglobulins 3.2.1 Resting Serum and Salivary lmrnunoglobulins in Athletes and Nonathletes 3.2.2 Exercise Training Effects on the Acute Exercise Immunoglobulin Response 3.2.3 Exercise Training Effects on Antibody Production 3.2.4 Changes in Serum lmrnunoglobulins With Moderate Exercise Training 4. Mechanisms and Future Research Needs 4.1 Acute Short Term Exercise 4.2 Acute Prolonged Endurance Exercise 4.3 Antibody Production in Response to Acute and Chronic Exercise 4.3.1 Effect of Hormones 4.3.2 Effect of T cells 4.3.3 Effect of lnterleukins 1 and 2 4.3.4 Effect of Psychological Stress 4.4 Clinical Significance

Summary

The effects of acute exercise (both graded-maximal and submaximal) and exercise training on resting immunoglobulin levels and immunoglobulin production are reviewed. Brief graded-maximal or intensive short term submaximal exercise tends to be associated with increases in serum immunoglobulins. the pattern of which does not vary between athletes and nonathletes. Plasma volume changes appear to largely explain these acute increases. Acute moderate exercise. such as a 45-minute bout of walking, on the other hand. has been associated with a transient rise in serum immunoglobulin levels despite no change in plasma volume. This increase is probably the result of contributions from extravascular protein pools and an increased lymph flow.

Sports Medicine I I (3) 1991

184

Total serum immunoglobulin changes following less than 40km of running are minor and/or statistically insignificant, although the concentration of IgG is observed to be at its lowest by 1.5 hours after exercise. The greatest effect of acute submaximal exercise appears to be on serum IgM levels which tend to increase, although results are somewhat inconsistent. Various mechanisms of stimulation have been proposed to explain the exercise-induced effect on IgM, which is the first antibody class produced in an immune response. These mechanisms include nonspecific noradrenergic sympathetic neural interactions with the immune system and the possibility of antigen stimulation through greater-than-normal quantities of microorganisms entering the body through both increased ventilation rates and breakdown of natural mucosal immunity by drying of airway secretions. When athletes run 45 to 75km at high intensities, serum immunoglobulin levels have been reported to be depressed for up to 2 days. Thus intense ultramarathon running may lead to greater and longer lasting decreases in serum immunoglobuli~llevels than following exercise of shorter duration. IgA and IgG. immunoglobulins commonly found in airway and alveolar space secretions, may have diffused from the serum during recovery from prolonged endurance exercise nonspecifically and/or in response to microbial agents and antigens introduced into the airways during the exercise bout. It has been well established that prolonged endurance exercise is associated with muscle cell damage and local inflammation. It has been hypothesised that natural (IgM) autoantibodies may be used to assist macrophages in disposal of muscle cell breakdown products. This could occur either by IgM binding to breakdown products present in the blood. followed by their clearance from the circulation, or it is possible that these antibodies may leave the circulation to carry out this same function in tissues. This reaction may need to be tempered by autoimmunosuppressive manoeuvres by the immune network to prevent immunoglobulin class switching to production of potentially harmful antibodies (IgG). Cross-sectional studies suggest that resting serum immunoglobulin levels in athletes and nonathletes are similar, especially when adjusted for plasma volume. However, some elite athletes, especially during the competitive season, may experience low concentrations of serum and secretory immunoglobulins, which have been described by some researchers as important factors leading to increased risk of infection. Muscular injections of immunoglobulins have proven useful in reducing the duration but not the incidence of infections in these competitive athletes. Moderate exercise training (45 minutes brisk walking, 5 sessions per week). on the other hand, has been associated in one study with a net 20% increase in serum immunoglobulins in comparison to a control group during a 15-week period, a change which was inversely correlated with total acute upper respiratory tract symptom days. The data overall suggest that while the combined psychosocial-physiological stress of competitive exercise training may be associated with depressed serum immunoglobulin levels in some elite athletes. moderate exercise training may lead to slightly improved serum immunoglobulin levels with contrasting effects on risk of infection. The data should still be regarded as preliminary, however. until confirmed with studies using more exacting methodologies. Following acute prolonged endurance exercise or several weeks of intensive exercise training, in vivo or in vitro antibody production has not been found to be altered. The effect of hormones, T cells. and interleukins I (IL-I) and 2 (IL-2) on antibody production, in response to endurance exercise have also been investigated. Although IL-I, a macrophage product, tends to be elevated in response to exercise, a corresponding increase in IL-2 has not been reported, perhaps because T cell function is often reported to be suppressed. Thus, following acute and chronic exercise, B cell growth and differentiation does not appear to be enhanced by soluble factors (including IL2) from the T helper cells.

1. Structure and Function of Immunoglobulins The major function of the immune system is

production of soluble and cellular components that provide immunity and protection against microbial infection. The major soluble components responsible for humoral immunity are the immu-

Exercise and immunoglobulins

.

185

noglobulins, a group of glycoprotein molecules that carry antibody activity, i.e. the property of combining specifically with an antigen (Goodman 1987; Turner 1989). Antibodies are produced by plasma cells, end-stage B lymphocytes, in response to foreign substances introduced into the body. The primary signal for antibody production varies greatly, depending on the initiating antigen. There is often a series of cooperative events required between macrophages (antigen presenting cells), T cells and B cells. These interactions are often under tight MHCII control and involve both interleukin (1L)I and IL-2. At the population expansion stage 1L-3, IL-4, IL-5, IL-6 and IL-7 are also important. B cell activation may take place in the absence of T cells, but the full development of B cells takes place only in the presence of T cells and their products (see fig. I). Basic antibody structure and function have been described in detail by several authors (Benjamini

& Leskowitz 1988; Burton & Gregory 1986; Goodman 1987). All antibodies are based on a monomer consisting of 2 long (heavy) and 2 short (light) polypeptide chains. This basic monomer can be chemically divided into 3 functional units. Two units, the Fab fragments, exhibit antigen binding activity, while the third unit, the Fc fragment (fragment crystalline), is involved in binding to molecules such as complement and to receptors on cells such as macrophages, neutrophils and lymphocytes that will trigger further host defense systems. The terminally situated Fc region of immunoglobulins binds to receptors found on the above cell types which then become associated with a variety of functions, including phagocytosis and antibodydependent cellular cytotoxicity. The major role of antibody is to combine with antigens to form free immune complexes which are then rapidly removed from the circulation or to combine with tissue, triggering various other effector functions such

B cell growth factors I L . ~I, L - ~IL-5, , IL-6. IL-7.1FN7

,,....,............

B-cell differentiation factors

Phagocytosis

(especially IgG1, IgG3)

Complement activetion (especially IgM, IgGi, lgG3)

Fig. 1. B cell activation and differentiation with T cell help. and immunoglobulin's role in phagocytosis and complement activation. IL = interleukin. IFNy = interferon.

Sports Medicine I I (3) 1991

186

as complement activation and phagocytosis (see fig. 1). Approximately 20% of total plasma proteins are immunoglobulins. There are 5 classes of antibodies or immunoglobulins termed immunoglobulin M (IgM), IgG, IgA, IgD and IgE. This review considers only IgM, IgG and IgA, the 3 immunoglobulins for which sufficient data relating to exercise physiology exists to draw some conclusions. IgM represents about 10% of total serum immunoglobulin (reference range in normal adults, 0.5 to 2.0 g/L) and is found primarily in the intravascular pool. IgM forms a pentameric structure (5 linked monomers) and is the predominant antibody produced early in an immune response, usually followed by IgG production. IgM is also the major immunoglobulin expressed on the surface of B cells (Lydyard 1989) and is an efficient complement-fixing immunoglobulin. The major antibody class in the serum is IgG, forming about 75% of the total immunoglobulin (5.4 to 16.lg/L). Four IgG subclasses (IgGI to IgG4) exist, which differ in their heavy chains. IgG is evenly distributed between intra- and extravascular pools. IgA forms about 15% of total serum immunoglobulin (0.8 to 3.0g/L), with 2 subclasses (IgAl and IgA2). Secretory IgA is the major antibody in various body secretions such as saliva, nasal secretions, sweat, colostrum and breast milk. Following primary challenge with antigen, there is a short initial lag phase in which no serum antibody can be detected (Widmann 1989). This phase is followed by a rapid increase principally in IgM, followed somewhat later by a switch to IgG production. A second antigenic challenge normally results in a rapid increase in antibody levels mainly of the IgG class with some IgM and IgA also produced. This rapid response to a second challenge results from the activation of specific memory cells formed during the primary response. Secretory IgA is the principal immunoglobulin in the local secretions coating the mucosal surfaces of the nasopharynx, oropharynx, conducting airways, eyes, gut and bladder (Reynolds 1987, 1988; Welliver & Ogra 1988). Salivary IgA levels have been reported to be very low in infection-prone

children compared with healthy controls (Lehtonen et al. 1987). IgG is the most common immunoglobulin found in airway and alveolar space secretions, diffusing into the lungs readily from the blood (Reynolds 1988). An important function of IgG in the lower respiratory tract is its specific antibody activity against microbial agents or antigens. With an absolute or functional deficiency of respiratory tract IgG, recurrent and chronic types of infections occur. Resistance to infection is due, in part to the presence of sufficient levels of serum and secretory immunoglobulins, especially the antigen-specific antibodies IgG and IgA. Serum immunoglobulin levels rise during the convalescent phase of an illness (Shimokata et al. 1988), and resistance to reinfection has generally been attributed to the presence of adequate levels of antigen-specific immunogloblins in serum or respiratory secretions (Welliver & Ogra 1988). 1. I B Cell Activation

The ability of the immune system to recognise antigens depends on the antibodies generated by the B cells and on the antigen receptors expressed by T cells. Each lymphocyte is only capable of recognising one of the many thousands of antigens present in the environment. An adequate response to an antigen depends on clonal selection whereby antigen-specific cells proliferate to mount an adequate immune response (Cooper 1987; Levitt & Cooper 1987; Male & Roitt 1989). After antigen stimulation, B lymphocytes can differentiate into plasma cells that secrete large amounts of immunoglobulin, or they can divide and return to a resting state as small memory B cells (see figure 1). Memory B cells are highly mobile, leaving the lymph node, spleen or other lymphoid tissue and joining the recirculating pool of lymphocytes (Nossal 1987). From the circulation, lymphocytes enter and patrol the tissues, or alternatively, re-enter the lymph nodes. Memory B cells can rapidly differentiate into plasma cells following a second exposure to the same antigen.

Exercise and Immunoglobulins

Antibody molecules are deployed throughout the circulation and the lymph. Since some antibodyforming cells are very long-lived, the potential for specific antibody production can persist for many months or years. In summary, when a foreign or infectious substance enters the body, the immune system is designed to respond in many different ways simultaneously so as to eliminate the invader rapidly. From the humoral viewpoint, when antigen is seen by naive cells (primary, nonsensitised immune cells), the first antibody induced is IgM; production of other classes of immunoglobulin may follow if the substance persists in the body. Simultaneously, if a clone of immune cells exist that has already seen the foreign substance, the presensitised memory cells are also activated. The daughter cells immediately synthesise and secrete antigenspecific antibody which is usually not IgM. Furthermore, since B cells can be stimulated by T cellproduced lymphokines and/or the binding of antigen to their receptors, either antigen or lymphokine can cause the induction of antibody synthesis. T cells play a crucial role in modulating the development of immunoglobulin-secreting plasma cells from immunoglobulin-bearing B cells, with T helper (CD4+) cells aiding and T suppressor (CD8+) cells inhibiting or controlling this process (Stobo 1987). The activated T helper cell produces several soluble factors including interleukin 2, y interferon, B cell growth factors, and B cell differentiation factors (see fig. l). These soluble factors bind to specific receptors on the activated B cell to give signals for growth and differentiation. A variety of other stimuli appear to modulate B cell activation including hormones, interleukin 1 (a macrophage product), and neuropeptides. These will be discussed further in the mechanistic section of this article.

2. Measurement of Serum Immunoglobulins Immunoglobulins can be assayed by a wide range of immunochemical techniques (Stites & Rodgers 1987; White & Ward 1986). Classically,

single radial immunodiffusion remains the universally accepted reference method for the measurement of serum IgM, IgG, and IgA. In this assay, monospecific antibody to the appropriate protein analyte (in this situation the heavy chain of the immunoglobulin acts as the antigen) is uniformly dispersed throughout an agar or agarose gel. A measured volume of serum or calibrant is placed in wells cut in the gel and allowed to diffuse radially in a moist atmosphere. As the antigen (immunoglobulin) diffuses into the antibody-containing gel, the antigen-antibody complexes precipitate at the point of equivalence. A ring of precipitation is formed, the area of which is proportional to the concentration of antigen. The assay is time-consuming, taking 2 to 3 days for IgG and 6 to 7 days for IgM and relatively insensitive to small changes in Ig concentrations. Electroimmunodiffusion speeds the migration of antigen by creating an electrophoretic field in the agar. Automated nephelometry techniques are also commonly used by investigators. Nephelometry is the measurement of light scattered at an angle from the beam path by particles in a clear medium (Stites & Rodgers 1987). The optimum wavelength is selected according to the size of the particles to be measured. The degree of light scatter is then proportional to the antigen concentration (provided antibody concentrations are maintained at a constant level). Immunoprecipitation, laser nephelometry and reaction rate nephelometry are other methods currently available (White & Ward 1986). Other methods include turbidimetry, immunofluormetric assays, radioimmunoassays and enzyme-linked immunosorbent assays (ELISA). ELISA is currently one of the most sensitive assays available for measuring a wide range of Ag/Ab reactions. It is commonly used for the measurement of specific antibodies to infectious agents and the quantitation of IgA, IgG subclasses and IgE.

3. Effects of Exercise on Immunoglobulins Relatively few published reports are available describing exercise-induced changes on serum and/ or salivary immunoglobulin levels. Many of these

Sports Medicine 11 (3) 1991

188

studies have utilised small numbers of subjects, often without control groups, weakening the power of statistical analyses and the ability to draw definitive conclusions. In this section, the reported effects of acute exercise (both graded-maximal and submaximal) on immunoglobulin levels are summarked first, followed by a review of the effects of exercise training on resting and acute-exercise responses. A separate section near the end of this article is reserved for describing probable mechanisms involved in these exercise-induced changes. 3.1 Effects of Acute Exercise on Immunoglobulins

3.1.1 Acute Graded-Maximal Exercise A few researchers have reported measurements of serum and salivary immunoglobulin levels before and immediately following graded-maximal exercise. Stephenson et al. (1985) exercised 5 healthy males maximally on a treadmill, and reported a significant 6% increase in IgG, and nonsignificant increases in IgA (13%) and IgM (1 1%). They reasoned that these changes could be accounted for by the 11.2% average decrease in plasma volume following maximal exercise. Poortmans (1970), however, reported that IgG and IgA increased significantly 2 minutes following maxi-

mal cycle exercise (1 1.8% and 14.4% respectively) in 28 Olympic athletes, and that this increase was slightly more than could be accounted for by the decrease in plasma volume. However, comparisons between 2 studies are hindered by the fact that the methods used in measuring plasma volume changes were different. Schouten et al. (1988) measured changes in salivary IgA in response to graded-maximal exercise in 84 men and 91 women. Salivary IgA decreased 10% in women but increased 10% in men in response to the test. Thus, the change in salivary IgA levels with graded-maximal exercise may be gender-specific.

3.1.2 Acute Intense Submaximal Exercise The literature reporting serum immunoglobulin changes in response to acute intense submaximal exercise is conflicting (see table I). Eberhardt (197 1) exercised 7 subjects for 20 minutes on a cycle ergometer at an intense workload of 1300 kgm/min, and compared immediate postexercise values with baseline. IgG stayed essentially the same, while IgA and IgM increased 23% and 43%, respectively. These changes were not statistically significant, however, and changes in plasma volume were not reported. Hanson and Flaherty (198 1) reported that serum immunoglobulins in 6 male runners who ran

Table I. Serum immunoglobulin changes in response to submaximal exercise Reference

No. of subjects

Less than 40km exertion Eberhardt (1971) 7

Mode of exercise

Hanson (1981)

6

Cycle ergometry Running

Nieman

10

Running

Greater than 40km exertion Israel (1982) 10 Israel (1982) 10 Poortmans (1979) 7

Running Running Running

Abbreviations: NC no reported change; ' p

< 0.05

Change in plasma volume

NC

NC vs baseline

Workload/ distance

% change from

% change from

baseline at 15 min

baseline at 18-25h

IgM

IgG

IgA

IgM

IgG

IgA

1300 kgm/min. 20 min 12.8km, 72% VOzmax 37.2km, 70% VOzmax

43

-6

23

34

2

2

29

-1

-5

5

3

0.4

1

1

2

45km 75km 1OOkm

-28 -20

-1 -4 7'

-21' -17 1

-24 -10

-18' -22' -4

-20 -28 -12


189

Table II. Mean ( 2 SE) serum immunoglobulin levels before, during and after a 3-hour treadmill run in 10 male marathon runners. P value represents significance of within-subject effects, repeated measures ANOVA. Variable (g/L)

'p

Baseline

Exercise (1 hour)

Recovery 1 (5 min)

Recovery 2 (1.5h)

Recovery 3 (6h)

Recovery 4 (21h)

P-value

< 0.05 vs baseline

13km at 72% 'Sro2rnax did not rise significantly above their baseline levels 10 minutes after exercise. IgM did rise 34%, and was still elevated by 29% the next morning, but these changes were statistically nonsignificant. Again, changes in plasma volume were not reported. We measured changes in serum immunoglobulins in 10 experienced marathoners who were exercised for 3 hours at marathon race pace in our laboratory (see table 11). [Although other results from this study have been published elsewhere (Berk et al. 1990; Nieman et al. 1989a), the serum immunoglobulin data from this study are being presented here for the first time.] Blood samples were collected 10 minutes prior to exercise (after 30 minutes of rest) after I hour of exercise, and after 5 minutes, 1.5, 6 and 2 1 hours of recovery. Levels of serum immunoglobulins, cortisol, adrenaline (epinephrine) and noradrenaline (norepinephrine) were measured. IgM (p = 0.09) and IgG (p = 0.07), but not IgA (p = 0.28), tended to show within-subject effects over time, using repeated measures ANOVA, despite no change in plasma volume (p = 0.78). IgG tended to decrease during recovery, reaching its lowest point at 1.5 hours after exercise (down 7.6% vs baseline, p = 0.13), rising slowly to baseline level 21 hours after exercise. Mean serum IgA values followed a similar pattern, but variance between subjects was greater, weakening statistical significance. IgM rose 7.2% during exercise (p < 0.05) and then fell to baseline levels throughout the recovery period. None of these changes were significantly correlated with changes in cortisol or catecholamines.

Israel et al. (1982) have reported changes in serum immunoglobulins in athletes running 45 or 75km at high intensity (see table I). Serum immunoglobulins were decreased 10 to 28% by 24 hours of recovery time, stayed low for 24 additional hours, and then rose slowly to baseline levels by the fourth recovery day. However, because of the large variance between subjects, many of the changes during recovery were not statistically significant. Poortmans and Haralambie (1979) reported a significant 7% increase in IgG in 11 male runners 15 minutes after completing a 1OOkm run despite no mean change in plasma volume. After 18 to 25 hours of recovery, serum IgG and IgA levels were 4% and 12%, respectively, below baseline levels. IgM was not measured. These data suggest that intense ultramarathon exertion may be associated with a greater and more prolonged depression of serum immunoglobulins than is sustained at shorter distances. Tomasi et al. (1982) have reported that resting salivary IgA levels in elite cross country skiers are lower than in age-matched controls, and that IgA levels decrease further after 2 to 3 hours of exhaustive ski competition. In a follow-up study on 8 well-trained male competitive bicyclists, each subject exercised on a bicycle ergometer for 2 hours at 70 to 75% 'Sro2,,, (MacKinnon et al. 1987). In contrast to the first study, resting salivary IgA levels were not found to be different from untrained control subjects. However, the skiers were studied during the winter in New York, the cyclists during the summer in New Mexico. Thus factors other than exercise training may be effecting resting sal-

ivary IgA levels. Immediately after the 2 hours of exercise, IgA from salivary, but not nasal, lavage fluid decreased an average of 65%, returning to baseline levels sometime between 1 and 24 hours after exercise. Serum IgM, IgG, IgA, and serum antibody titres to specific antigens were unchanged after exercise. In addition, in vitro production of IgG and IgA by pokeweed-stimulated peripheral blood lymphocytes was not altered following exercise. Other researchers have also investigated the ability of B lymphocytes to produce antibodies following exertion. Tvede et al. (1989) cycled 20 subjects at 80% V o 2 m a x for 1 hour, and in contrast to Mackinnon et al. (1987), reported that B lymphocyte function was suppressed for 2 hours of recovery. This decrease in B lymphocyte function appeared to be due to an inhibitory effect from monocytes activated during the exercise bout. Eskola et al. (1978) reported that 4 well-trained runners immunised 30 minutes after a marathon race showed no impairment in ability to produce antibody (in vivo, after 14 days) to tetanus toxoid vaccination. This occurred despite a significant decrease in T lymphocyte transformation (in response to in vitro mitogen stimulation with concanavalin A) which persisted for several hours following the marathon event. Although activated T cells are important in helping B cells develop into immunoglobulin-secreting plasma cells, the transient nature of the decrease in T lymphocyte transformation following prolonged endurance exercise may mean that the longer term in vivo antibody response is unaltered. To summarise, brief graded-maximal or heavy short term submaximal exercise tends to be associated with increases in serum immunoglobulins. Serum immunoglobulin changes following less than 40km of running are minor and/or statistically insignificant, although there is a tendency for IgG to decrease to its lowest levels 1.5 hours after exercise. The greatest effect of acute prolonged exercise appers to be on IgM which tends to increase, although results are somewhat inconsistent. In contrast, when athletes run greater than 40km, serum immunoglobulin levels tend to be depressed for up

to 2 days of recovery. Thus ultramarathon running may lead to greater and longer lasting decreases in serum immunoglobulin levels than exercise of shorter duration. Finally, long endurance exercise has been associated in 2 studies with significantly lower salivary IgA levels. 3.1.3 Acute Moderate Submaximal Exercise We recently examined the extent and duration of changes in serum immunoglobulin levels in 12 women who walked 45 minutes at 60% '?02,,, in a laboratory setting (Nehlsen-Cannarella et al. 1991b). Each subject reported to the laboratory in a 12-hour fasted condition at 06.30 on 2 days, 1 week apart. The first day, half the group was randomly selected to engage in exercise, the other half to rest, with conditions reversed the following week. Care was taken to ensure that measurement procedures and the laboratory environment were exactly the same for both conditions. Blood samples were taken at 07.00, 08.00,09.30, 1 1.00, 13.00, and the following morning at 07.00. During the experimental session, subjects walked at 60% 002,,, for 45 minutes from 07.15 to 08.00 while subjects in the rest condition sat quietly nearby. The patterns of change for serum IgG, IgA, and IgM were significantly different (p = 0.001, p < 0.001, p = 0.010, respectively) between conditions despite no change in plasma volume (table 111). IgG rose 7.2% immediately following exercise, and then returned to baseline 1.5 hours later, which contrasted significantly with changes in the rest condition. These same patterns of change occurred also with IgA and IgM. Thus walking, as opposed to other forms of exercise, may lead to a transient rise in serum immunoglobulins. Mechanisms underlying this increase are discussed later in this article. 3.2 Effects of Exercise Training on Serum Immunoglobulins 3.2.1 Resting Serum and Salivary Immunoglobulins in Athletes and Nonathletes Most reserchers have reported that trained athletes have resting serum immunoglobulin levels within the normal reference range and similar to


191

Table Ill. Effect of acute moderate exercise on serum immunoglobulin levels (mean + SE) in 12 women (Nehlsen-Cannarella et al. 1991 b). Blood samples were taken at 07.00, 08.00, 09.30, 11.00, 13.00, and the following morning at 07.00. During the experimental for 45 minutes from 07.15 to 08.00 while subjects in the rest condition sat quietly nearby session, subjects walked at 60% VO2,, Condition

immunoglobulin level (g/L) baseline (07.00)

recovery 1 (08.00)

recovery 2 (09.30)

recovery 3 (1 1.OO)

recovery 4 (13.00)

recovery 5 (07.00)

Effect (condition X time) p-value

IgG Exercise Rest IgA Exercise Rest IgM Exercise Rest Significant differences between conditions in change from baseline: ' p

those of sedentary controls (Green et al. 1981; Hanson & Flaherty 1981; Haralambie & Keul 1970; Israel et al. 1982; Nieman et al. 1989b; Poortmans 1970). Green et al. (1981), for example, reported that mean serum IgM, lgG, and IgA values in marathon runners did not differ significantly from sedentary controls. In addition, there were no significant differences in immunoglobulin values when higher and lower mileage groups were compared. Petrova et al. (1983) of the Soviet Union, however, have reported that serum and secretory immunoglobulin levels decreased in response to increasing amounts of exercise stress in 9 1 sportsmen. Von Weiss et al. (1985) of the Federal Republic of Germany in a large study of 950 (310 female, 640 male) athletes engaging in any of 43 different sports reported that 7.3% had serum IgA levels below 0.9 g/L in comparison to 0.2% of nonathletes. In addition, 5.4% and 2.0% of these athletes were below the normal range for IgG and IgM, respectively. Mean (range) serum IgM, IgG and IgA values for these athletes were 1.5 (0.3 to 4.5). 12.2 (3.2 to 23. I), and 1.79 (0.3 to 7.0) g/L, respectively. No relationship between serum IgA and training variables (aerobic capacity, number of years, frequency, duration, and intensity of training) was found, however. Rijcker et al. (1976) have reported significantly

< 0.05;

"

p

< 0.01.

lower serum IgM and IgG, but not IgA, levels (g/ L) in 40 elite endurance athletes in comparison to 49 sedentary controls. However, with adjustment for the expanded plasma volumes of the athletes (600ml greater), serum IgM and IgG levels were not found to be significantly different between these 2 groups. Thus some of the disparity in results between the studies reviewed in this section may be due to the relationship between exercise training and an expanded plasma volume. Nonetheless, researchers have reported that low serum immunoglobulin levels in some elite athletes is accompanied by an increased susceptibility to bacterial and viral infections (Ricken & Kindermann 1986). Intramuscular administration of 7 globulin has proven useful in helping these athletes avoid acute respiratory tract infections during foreign tours and competition (Frolich et al. 1987; Ricken & Kinderman 1986). Frohlich et al. (1%7) in a double-blind placebo-controlled sfudy of 20 German Armed Forces swimmers reported that intramuscular injections of 7 globulin (l.6g initial dose, then 0.8g every 4 weeks for 12 weeks) resulted in a 3-fold reduction in the duration (but not the incidence) of infections. Earlier in this article we summarised 2 studies showing that while elite crosscountry skiers during the winter season have resting salivary IgA that are


significantly lower than controls (Tomasi et al. 1982), well-trained competitive bicyclists during the summer season do not have values different from controls (MacKinnon et al. 1987). In a study of 175 healthy young adults no relationship between varying levels of physical activity and salivary IgA was demonstrated (Schouten et al. 1988). Thus, the effect of exercise training on resting salivary IgA levels, recorded so far, may be related more to the temperature of the air ventilated than differences in training workload. The results of the studies reviewed in this section may also be confounded by the types of subjects used. Exercise training may be associated with little change in resting serum and salivary immunoglobulins until extremely intensive training is repetitively engaged in by athletes. In addition, elite athletes may experience a greater likelihood of decreased immunoglobulin levels during the competitive season due to the combined psychosocial and physiological stress. Psychosocial stress alone has been associated with decreased salivary IgA levels (Jemmott & Locke 1984) [see also section 4.3.41. 3.2.2 Exercise Training Effects on the Acute Exercise Immunoglobulin Response Do athletes differ from nonathletes in their serum immunoglobulin response to acute gradedmaximal exercise? In a study conducted in our laboratories, I I male marathon runners were recruited from a local running club and compared with 9 sedentary, age-matched controls (Nieman et al. 1989b). Although each subject was experienced, having run at least 3 marathons within the previous 5 years, these athletes were not elite performers. During the 3 previous years, the subjects had averaged 68.4 + 6.3 kmlweek of running, with an average marathon time of 3.1 + 0.1 hours. Venous blood samples were obtained before (average of 3 baseline samples), during (every 5 minutes), and following graded-maximal exercise (5 samples during 45 minutes of recovery) utilising the Balke protocol. Average baseline, exercise, and recovery values for IgM, IgG and IgA were not significantly dif-

ferent between athletes and nonathletes (see fig. 2). Patterns of increase during exercise and decrease during recovery for all parameters did not differ significantly between groups. MANOVA for increase in parameter values from baseline to the first recovery value (5 minutes after exercise) showed significant increases for IgM and IgG (p < 0.01) with a trend for increase in IgA (p < 0.06). For all subjects combined, the average increases from baseline to 5 minutes after exercise for IgM, IgG, and IgA were 17.7, 16.7, and 12.2%, respectively. Parameter values unadjusted for plasma volume changes were not significantly elevated above baseline values 30 minutes after maximal exercise. Plasma volume in athletes and nonathletes 15 minutes after exercise were 8.5 + 4.5 and 8.1 -+ 0.8% lower, respectively, than baseline values. When 15-minute recovery serum immunoglobulin values for the entire group were adjusted for this plasma volume decrease, parameter values were not significantly elevated above baseline values. Furthermore, there were no significant relationships between baseline and 5-minute recovery values of serum immunoglobulins, and cortisol, adrenaline or noradrenaline using hormone values as covariates, although there was a trend for adrenaline values to be associated with IgG (p < 0.06) The major conclusion of this study is that athletes and nonathletes tend to show the same pattern of increase in serum immunoglobulins in response to graded-maximal exercise which is best explained by the concomitant decrease in plasma volume.

3.2.3 Exercise Training Effects on Antibody Production We are aware of 3 published reports on the effect of exercise training on antibody production in vivo. In all 3 studies, the primary antibody response to antigen was not affected by exercise. Davis et al. (1986) showed that 10 weeks of training (running 1 houriday, 5 daysiweek) did not enhance or decrease the antibody responses of 4 groups of Swiss Webster mice to the influenza virus vaccine. This was confirmed in another study where antibody production in response to a pri-


193

Athletes IgG IgA IgM

Nonathletes W

,

A

A

0

.1

I

q

I

I7

7 I.

.!I7![

I

I

10

Baseline

I

I I

I

I

20

40

60

80

I

100

I

I

I

I

I

45 min recovery

I

Exercise (%'&rnax) Fig. 2. Serum immunoglobulin changes in ath!etes and nonathletes in response to graded maximal exercise (Balke protocal).

Exercise results are expressed in relation to %V02,,,. Nieman et al. (1989b).

Blood samples were collected every 5 minutes of exercise. Data from

mary antigen inoculation in exercise-trained rats and mice was found not to be significantly different to that of nonexercised animals (Keast et al. 1988). Douglass (1974), however, showed that while primary responses to an antigen (purified diphtheria toxoid) were not altered by daily exercising (2 hours per day swimming for 11 weeks) in mice, secondary responses were enhanced. 3.2.4 Changes in Serum Immunoglobulins with Moderate Exercise Training To our knowledge, only one randomised controlled exercise training study, using human subjects, has evaluated the effects of exercise on serum immunoglobulins (Nehlsen-Cannarella et al. 1991a). The relationship between moderate exercise training (five 45-minute sessions per week, brisk walking at 60% of heart rate reserve for 15 weeks) and changes in serum immunoglobulins was investigated in a group of 36 sedentary, mildly obese women. The study was conducted using a 2 (exercise and nonexercise groups) x 3 (baseline, 6-, 15-

week testing sessions) factorial design, with data analysed using repeated measures ANOVA. The pattern of change over time between groups for serum IgM, IgG and IgA levels was significantly different (figs 3 and 4). The exercise group experienced significantly greater increases than the nonexercise group for each of the 3 serum immunoglobulins from baseline to 15 weeks. The average percentage increases from baseline to 15 weeks in serum immunoglobulins for exercise vs nonexercise groups were: IgM, 22.8 f 5.0 vs 2.7 f 5.4%; IgG, 24.1 + 4.2 vs 5.3 f 7.3%; IgA, 31.7 f 4.5 vs 12.9 + 7.8%, respectively. Using data from daily logs kept by each subject, the exercise group reported significantly fewer days with symptoms of acute upper respiratory tract infections (URI) than did the nonexercise group during the entire 15-week study (5.1 f 1.2 vs 10.8 f 2.3 days, respectively, p 6 0.039). Improvement in submaximal cardiorespiratory fitness, and increase in serum immunoglobulins were both correlated significantly with a reduction in total symptom days

Sports Mcdicinc I I (3) 1991

El Exerclse

'41

8 Nonexerc~se

Weeks of training

Fig. 3. The effects of modcratc cxcrcisc training on the concentration of scrum IgG. Data from Nehlscn-Cannarclla el al. (1991a). * p < 0.05 1,s basclinc. within-group cffccl.

with URI (see fig. 5). Improvement in submaximal cardiorespiratory fitness was also correlated significantly with increases in each of the 3 serum immunoglobulins. These results need to be confirmed and extended by others, to better define the association between serum immunoglobulin levels and the apparent protective effect against URI of moderate exercise training. Liesen et al. (1976) reported that 1 1 weeks of endurance training in 22 elderly male subjects (aged 55 to 70) was associated with significantly higher serum immunoglobulin levels. However, the study design did not include a randomised control group. In the study by NehlsenCannarella et al. (1991a), the net exercise traininginduced changes were approximately 20% for each serum immunoglobulin. Although these changes are rather small and within the range of normal values, the significant findings of inverse correlations with URI suggest that even small changes in serum immunoglobulins may have important clinical applications. In summary, cross-sectional studies suggest that serum immunoglobulin levels between athletes and nonathletes at rest and during graded-maximal exercise are similar, especially when adjusted for plasma volume. However, some elite athletes, es-

pecially during the competitive season, may experience low serum immunoglobulin levels leading to increased risk of infection. In contrast, moderate exercise training has been associated in one study with a net 20% increase in serum immunoglobulins in comparison to a control group during a 15-week period, a change which was inversely correlated with total URI symptom days. These data suggest that while the combined psychosocial-physiological stress of competitive exercise training may be associated with depressed serum immunoglobulin levels in some elite athletes, moderate exercise training may lead to slightly improved serum immunoglobulin levels. Further research is needed to better define the relationship between degree of exercise training and serum and secretory immunoglobulins.

4. Mechanisms and Future Research Needs 4.1 Acute Short Term Exercise We would agree with Stephenson et al. (1985) that plasma volume changes appear to largely explain much of the acute increase seen in immuIgA Exerclse Nonexerc~se

8

.

IgM

Weeks of training Fig. 4. The cnkcts of moderate exercise training on the concentrations of scrum IgA and IgM. Data from Nehlscn-Cannarella el al. (1991a). * p < 0.05 I:$ baseline, within-group cffcct.


Total URI symptom days 15-week period

Increase in r = -0.44, p = 0.008 (lgG) Improvement immunoglobulin r = -0.44. p = 0.006 (lgA) in fitness r = -0.38, p = 0.024 (lgM) levels

Fig. 5. Correlations between improvement in submaximal cardiorespiratory fitness (decrease in stage 2 heart rate), increase in serum immunoglobulins and total upper respiratory tract infection (URI) symptoms days during a 15-week moderate exercise training study. Data from Nehlsen-Cannarella et al. (1991a).

noglobulins following graded-maximal exercise. Poortmans (1 970), however, has reasoned that immediately following graded-maximal exercise, contributions to the circulating immunoglobulins may occur from rapidly exchangeable extravascular plasma protein pools. Following a 45-minute walking bout we recently showed that serum immunoglobulins were transiently elevated despite no change in plasma volume (Nehlsen-Cannarella 199 1b). Thus, acute exercise may increase the flow of lymph, which contains immunoglobulins, into the vascular compartment (Wells et al. 1982). The intra- and extravascular protein pools are in a state of dynamic equilibrium, and acute exercise may favour a slight increase in the total influx of various types of proteins, including globulins, into the intravascular pool. Thus at least a portion of the increase in serum immunoglobulins seen following acute short term exercise may be from extravascular pools and the lymph. Convincing data to support this supposition are lacking, however, and further research is needed to measure the effect of acute exercise on immunoglobulin retrafficking.

4.2 Acute Prolonged Endurance Exercise Acute prolonged endurance exercise of less than 40km of running is associated with minor changes in serum immunoglobulin levels. Changes in serum IgM, however, tend to be greater than for IgG and IgA (see table I). Data from table I1 show that serum IgM rose significantly during a 3-hour bout of running while serum IgG and IgA levels did not change. Serum IgG and IgA, however, tended to reach their lowest levels at 1.5 hours of recovery while igM stayed at baseline levels. Following high-intensity 45 or 75km runs, an even greater and longer lasting suppression of serum immunoglobulins was experienced (Israel et al. 1982). If endurance exercise in some way induces antibody synthesis, it is probable that IgM would increase more rapidly than IgG and IgA in the serum. IgM is the first antibody produced in an immune response, whether the immune system is responding on the primary or the secondary level. If future research can help resolve the issue as to whether or not long endurance exercise is associated with significantly increased serum IgM, the central issue then involves the mechanism(s) of stimulation of this immunoglobulin class. Exciting new insights into noradrenergic sympathetic neural interactions with the immune system have been reported (Felten et al. 1987; Madden et al. 1989) that underscore the complexity of immune regulation. In addition to these demonstrations of neural innervation of lymphoid tissues, it has been shown that lymphocytes have hormone receptors that are structurally and functionally similar to receptors of the neuroendocrine system. Furthermore, both the neuroendocrine and the immune system can produce the same, or similar ligands, capable of fitting these receptors. Measurements of various immune parameters in experimental animals exposed to differing degrees of environmental stress have demonstrated a range of responses, from suppression to elevation and back to suppression again (Weiss et al. 1989). The difficulty lying ahead for all aspects of this research lies in the interpretation and coordination of observations made on subjects exposed to dif-


fering stimuli at various levels of exercise in different situations. In normal whole-animal systems, multiple paracrine substances are available to interact to varying degrees under different conditions to form extensive soluble and cellular networks. These variables then allow for the generation of any one of a number of different outcomes designed to promote a balanced overall function in the whole animal. Such an interactive network which might develop as a result of exercise might involve the immune system receiving a signal from the neuroendocrine network that activates T lymphocytes t o secrete lymphokines. These soluble substances, in turn, would stimulate B lymphocytes to synthesise and secrete immunoglobulin. On the other hand, the mechanism of stimulation might be as simple as the synthesis and release of neuroleukin (a soluble protein mediating trophic communication across the nerve-muscle synapse) that is known to induce B cell differentiation to antibody-secreting cells (Gurney et al. 1986). Several other such factors are known to exist, such as the neuropeptides that are capable of exerting some regulation on the mucosal immune system (e.g. somatostatin, substance P and vasoactive intestinal peptide) [Stead et al. 19871. In addition, exercise-induced changes in hormones and varims components of both the innate and adaptive immune systems may allow infectious microbial agents to gain a temporary foothold in the body immediately following exercise (Berk et al. 1990; Lewicki et al. 1987; Nieman et al. 1989). Changes in serum immunoglobulin levels may reflect a reactive attempt by the adaptive immune system to combat growth of these infectious agents, with local variations in IgA and/or IgG being important in determining the final outcome of an infection (Reynolds 1987, 1988). It has been well established that severe exertion or heavy exercise training is associated with muscle cell damage and local inflammation (Armstrong 1986). Prolonged physical exercise induces an acute phase reaction for 1 to 4 days that is similar to that seen during a septic or aseptic inflammatory response (Dufaux et al. 1984; Liesen et al. 1977; Tay-

lor et al. 1987). An elevation of C-reactive protein (CRP) accompanies this response. CRP can activate the complement system and initiate cell-damaging and inflammatory reactions which are necessary for resolution and repair processes. Other plasma enzyme systems are also important in mediating the inflammatory response and the resolution of tissue damage. Furthermore, macrophage activity has been reported to be increased in the tissues of athletes following endurance running (Fehr et al. 1989). Macrophages have receptors on their surface which allow them to react nonspecifically to a variety of substances, with this process enhanced by the presence of opsonins (primarily complement and antibody) [Johnston 19881. Therefore, decreases in serum immunoglobulins which can occur following prolonged endurance running (Israel et al. 1982) may reflect involvement in the process of repair (Coleman et al. 1989). 4.3 Antibody Production in Response to Acute and Chronic Exercise In general, following acute prolonged endurance exercise (Eskola et al. 1978; MacKinnon et al. 1987), or several weeks of heavy exercise training using animal models (Davis et al. 1986; Douglass 1974; Keast et al. 1988), the in vivo or in vitro production of antibody has been found to be unaltered. In a recent well designed study by Tvede et al. (1989), however, P-lymphocyte function was reported to be suppressed for 2 hours following 1 hour of intensive cycling. Acute intensive exercise would be expected to affect P-lymphocyte function because a number of hormones and related compounds can influence the ability of B cells to produce immunoglobulins. 4.3.1 Efects of Hormones During endurance running at 70% 00~,,, for example, significant increases in catecholamines and cortisol occur (Berk et al. 1990). Adrenaline (epinephrine) is a potent adrenergic agonist, and Padrenergic stimulation of lymphocytes has been reported to produce several changes in the immune system including increased antibody production


(Crary et al. 1983). Noradrenaline (norepinephrine) has been shown in vitro to enhance lipopolysaccharide-induced B cell proliferation and maturation into IgG-, IgA- and IgM-secreting cells (Li et al. 1990). Sherman et al. (1973) have shown that lymphocyte cultures briefly exposed to physiological concentrations of hydrocortisone or adrenaline formed significantly increased quantities of immunoglobulins during the 72-hour incubation. Serum IgG and IgA and to a lesser extent IgM levels, on the other hand, have been reported to be suppressed in vivo following a brief course of daily corticosteroid administration (Cupps & Fauci 1982). Maximum suppression of the immunoglob'ulins occurred 2 to 4 weeks after treatment. An increased rate of immunoglobulin catabolism was noted during corticosteroid treatment, with a decreased rate of production occumng later. Data currently available suggest that if unusually heavy training leads to a stressful condition where cortisol is elevated, serum immunoglobulin levels may drop due to increased catabolism and decreased production which may, in part, explain the findings of Frijhlich et al. (1987), Petrova et al. (1983), Ricken & Kindermann (1986) and Von Weiss et al. (1985). More research is needed to test the responsiveness of B lymphocytes in vivo under variable hormone conditions known to develop following both acute and chronic exercise.

4.3.2 Effect of T Cells The regulatory functions of T cells are represented by their ability to amplify cell-mediated cytotoxicity by other T cells and immunoglobulin production by B cells (Stobo 1987) [fig. 11. As reviewed earlier, CD4+ T helper cells help B cells develop into immunoglobulin-secreting plasma cells, while CD8+ T suppressor cells inhibit this differentiation. Thus immunoglobulin production by B cells, in response to antigen or mitogen stimulation, is influenced by the CD4+/CD8+ratio (Crary et al. 1983). This may explain the contrast in the results of Hedfors et al. (1983) with Eskola et al. (1978) and MacKinnon et al. (1987). Hedfors et al. (1983) reported that in vitro antibody production (7 day in-

cubation of 1 million lymphocytes with and without pokeweed mitogen) was decreased after 10 minutes of moderate bicycling compared with antibody production of pre-exercise samples. The number of T helper cells in peripheral blood, decreased in these subjects, resulting in a lowered CD4+/CD8+ ratio. The decrease in this ratio has been reported by several researchers following both maximal and submaximal exercise of short duration (Berk et al. 1988; Brahmi et al. 1985; Landmann et al. 1984; Lewicki et al. 1987). Interestingly, the CD4+ cell count does not appear to change significantly during recovery from marathon running (Nieman et al. 1989a). In this study, the CD4+/CD8+ actually increased significantly (39%) during 2 1 hours of recovery from marathon running due to a significant decrease in the CD8+ cell count. Thus the effect of exercise on in vitro antibody production (unaltered after long endurance exercise but decreased after short duration exercise) may in part depend on concomitant changes in the CD4+/CD8+.

4.3.3 Effect of Interleukins I and 2 Macrophages and antigen presenting cells (APC) are primary sources of interleukin 1 (IG1). I G l stimulates, as part of its activity spectrum, CD4+ lymphocyte functions and enhances the production of IL-2 and B cell growth factor which then support B lymphocyte proliferation and antibody production (Johnston 1988). Following both graded-maximal (Lewicki.et al. 1988) and 1 hour of moderate cycle ergometry (Cannon et al. 1986), IL-I production in vitro and its activity have been reported to increase, peaking about 2 to 3 hours after exercise. The exercise-induced increase in I G 1 would be expected to lead to an increase in IL-2 production. However, IL-2 production in vitro has been found to be diminished following acute gradedmaximal exercise (Lewicki et al. 1988) and after 4.5 months of heavy swim exercise training in young rats (Pahlavani et al. 1988). Thus, although IL-I synthesis was increased in response to exercise, a corresponding increase in IL-2 was not found, suggesting that there may be no enhancement of antibody production from this route. This


conclusion is supported by Pahlavani et al. (1988) in their study of male rats. Lymphocytes isolated from the spleens of young exercise-trained rats (2 hours of swimming per day for 4.5 months) showed reduced 1L-2 production and proliferation in response to lipopolysaccharides, suggesting that antibody production might be suppressed in these animals. The failure of both 1L-l and 1L-2 levels to increase during exercise may be related to an exercise-induced failure of T lymphocytes to respond to mitogens. Proliferation rates of human T lymphocytes (to concanavalin A or phytohaemaglutinin mitogens) isolated after both acute exercise (Blecha & Minocha 1983; Eskola et al. 1978; Gmiinder et al. 1988; Hedfors et al. 1976; Mahan & Young 1989; Robertson et al. 1981) and chronic exercise challenge (Hoffman-Goetz et al. 1986) are generally found to be depressed. At present, the factors explaining this depression are not known, but may be due in part to inhibition by suppressor T cells (Oppenheim et al. 1987). Thus, following acute and chronic exercise challenge, B cell growth and differentiation is probably not enhanced by soluble factors (including IL-2) from the T helper cells. The net result is that antibody production may be unaltered or decreased in response to both acute and chronic exercise from this path.

4.3.4Effect of Psychological Stress Stress is associated with changes in various humoral and cellular immune mechanisms including suppressed antibody response in both laboratory animals and humans (Shavit et al. 1985). Several factors modulate the effect of stress on the immune system, however, including the chronicity, timing, intensity and controllability of the stressor (Gisler et a]. 1971; Shavit et al. 1985; Weiss et al. 1989). Jemmott & Locke (1984) have reviewed the studies evaluating the effect of stress on the immunoglobulins, and concluded that the secretion of IgA is lower during periods of high stress than low stress. In addition, @-lymphocyte transformation responses to mitogenic stimulation is diminished during periods of high stress. On the other hand, in a 15-week moderate exercise training study noted

earlier (Nehlsen-Cannarella et al. 1990), exercised subjects experienced a significant increase in general well-being in comparison to controls (Cramer et al. 1991 ). Therefore, psychological stress may confound the relationship between exercise and the dynamics ofimmunoglobulins, and should be considered as an important factor in such studies. 4.4 Clinical Significance There are several reports in the literature that athletes engaged in heavy training are at increased risk for acute respiratory infections (Jokl 1974; Nieman et al. 1988; Peters & Bateman 1983). This potential increase in susceptibility to infection could involve changes in many aspects of the immune system. Cortisol levels increase in response to the stress of the exercise workload, and have been demonstrated to be related to suppression of several components of the immune system (Berk et al. 1990; Cupps & Fauci 1982). Although this review has determined that serum and salivary immunoglobulins may be low in some athletes, especially elite athletes during the competitive season, data to link these low levels of immunoglobulins to increased acute respiratory infection is unconvincing (Israel et al. 1982). The effect of long endurance exercise on the concentration of salivary IgA would, however, appear to have some practical significance. The failure of researchers to find that nasal lavage fluid IgA levels were altered or that in vrtro or in vivo production of immunoglobulins was affected makes clinical application of the current information difficult (Drutz & Mills 1987; MacKinnon et al. 1987). Thus, at this time, the increased susceptibility of some endurance athletes to URI may be better explained by exercise-induced deficiencies of innate immunity, including lower serum complement levels (Nieman et al. 1989b), decreased phagocytic function (Lewicki et al. 1987; Petrova et al. 1983), and acute decreases in natural killer cell activity following prolonged exercise (Berk et al. 1990). than by alterations in adaptive immunity. The study by Frohlich et al. (1987), showing that muscular injections of 7 globulin were associated

Exercise a n d l m m u n o g l o b u l i n s

with a 3-fold reduction in the duration but not the incidence of infections. is very interesting. The clinical effects of this study are very similar to those of Nehlsen-Cannarella et al. (1990) who reported that 15 weeks of moderate exercise training led to a net 20% increase in serum immunoglobulins which was correlated with a 50% decrease in total symptoms days of acute respiratory tract infections in comparison to a sedentary control group. In this same study, the number of separate URI incidents did not vary between groups (1.0 + 0.2 vs 1.1 + 0.2 for exercise and nonexercise groups, respectively, p < 0.693) but the number of URI symptom days per incident was significantly lower (about half) in the exercise group (p < 0.049) [Nieman et al. 19901. Thus muscular immunoglobulin injections in athletes and moderate exercise training by nonathletes have similar effects, specifically in reducing the duration but not the incidence of URI.

Acknowledgements T h e a u t h o r s acknowledge t h e assistance of G u n t e r Reiss, D H S c , w h o translated t h e literature f r o m G e r m a n sports medicine journals.

References Armstrong RB. Muscle damage and endurance events. Sports Medicine 3: 370-381. 1986 Benjamini E, Leskowitz S. Immunology: a short course. Alan R. Liss. Inc.. New York. 1988 Berk LS, Tan SA. Nieman DC. Eby WC. Stress from maximal exercise modifies T-helper and T-suppressor lymphocyte subpopulations and their ratio in man. In Dotson CO (Ed) Exercise Physiology, current selected research. Vol. 3. AMS Press. Inc.. New York. 1988 Berk LS. Nieman DC. Youngberg WS. Arabatzis K. SimpsonWestberg M, et al. The effect of long endurance running on natural killer cells in marathoners. Medicine and Science in Sports and Exercise 22: 207-2 12, 1990 Blecha F. Minocha HC. Suppressed lymphocyte blastogenic responses and enhanced rn vilro growth of infectious bovine rhinotracheitis virus in stressed feeder calves. American Journal of Veterinarian Research 44: 2 145-2 148. 1983 Brahmi Z, Thomas JE, Park M, Park M. Dowdeswell IRG. The effect of acute exercise on natural killer-cell activity of trained and sedentary human subjects. Journal of Clinical Immunology 5: 321-328. 1985 Burton DR. Gregory L. Structure and function of immunoglobulins. In French MAH (Ed.) Immunoglobulins in health and disease, MTP Press Limited. Boston. 1986 Cannon JG. Evans WJ. Hughes VA, Meredith CN. Dinarello CA. Physiological mechanisms contributing to increased interleukin-l secretion. Journal of Applied Physiology 6 I: 1869- 1874. 1986

Cohen IR, Cooke A. Natural autoantibodies may prevent autoimmune disease. Immunology Today 7: 363-364. 1986 Coleman RM, Lombard MF. Sicard RE. Rencricca NJ. Fundamental immunology, Wm C. Brown Publishers. Dubuque. 1989 Cooper MD. Current concepts, B lymphocytes: normal development and function. New England Journal of Medicine 3 17: 1452-1456, 1987 Cramer SR, Nieman DC. Lee JW. The effect of moderate exercise training on psychological wellbeing and mood state in women. Journal of Psychosomatic Research, in press, 1991 Crary B, Hauser SL, Borysenko M, Kutz 1, Hoban C , et al. Epinephrine-induced changes in the distribution of lymphocyte subsets in peripheral blood of humans. Journal of Immunology 131: 1178-1 181, 1983 Cupps TR, Fauci AS. Corticosteroid-mediated immunoregulation in man. Immunological Review 65: 133-155, 1982 Davis MA, Kendrick ZV, Paolone M. The measurement of antibody response to influenza virus vaccine in aerobically trained mice. Medicine and Science in Sports and Exercise 18: S9, 1986 Douglass JH. The effects of physical training on the immunological response in mice. Journal of Sports Medicine 14: 48-54, 1974 Drutz DJ, Mills J. Immunity and infection. In Stites DP et al. (Eds) Basic & clinical immunology. 6th ed., Appleton & Lange, Los Altos, CA, 1987 Dufaux B, Order U, Geyer H, Hollmann W. C-Reactive protein serum concentrations in well-trained athletes. International Journal of Sports Medicine 5: 102-106. 1984 Eberhardt A. Influence of motor activity on some serologic mechanisms of nonspecific immunity of the organism. Acta Physiologica Polonica 22: 201-212, 1971 Eskola J, Ruuskanen 0 , Soppi E, Viljanen MK, Jarvinen M, et al. Effect of sport stress on lymphocyte transformation and antibody formation. Clinical and Experimental Immunology 32: 339-345, 1978 Fehr HG, Lotzerich H, Michna H. Human macrophage function and physical exercise: phagocytic and histochemical studies. European Journal of Applied Physiology 58: 6 13-6 1 7, 1989 Felten DJ, Felten SY, Bellinger DL, Carlson SL, Ackerman KD, et al. Noradrenergic sympathetic neural interactions with the immune system: structure and function. Immunological Review 100: 225-260, 1987 Frohlich J. Simon G, Schmidt A, Hitschhold T, Bierther M. Disposition to infections of athletes during treatment with immunoglobulins. International Journal of Sports Medicine 8: 119, 1987 Gisler RH. Bussard AE, Mazie JC, Hess R. Hormonal regulation of the immune response I. Induction of an immune response in vilro with lymphoid cells from mice exposed to acute systemic stress. Cell Immunology 2: 634-645, 1971 Gmunder FK, Lorenzi G, Bechler B, Joller P, Muller J, et al. Effect of long-term physical exercise on lymphocyte reactivity: similarity to spaceflight reactions. Aviation, Space and Environmental Medicine 59: 146-151, 1988 Golub ES. Immunology: a synthesis, Sinauer Associates, Inc., Sunderland, MA, 1987 Goodman JW. Immunoglobulins I: structure and function. In Stites DP et al. (Eds) Basic and clinical immunology, 6th ed., Appleton & Lange, Los Altos, CA, 1987 Green RL. Kaplan SS. Rabin BS, Stanitski CL, Zdziarski U. Immune function in marathon runners. Annals of Allergy 47: 7375. 1981 Gurney ME, Apatoff BR, Spear GT, Baumel MJ, Antel JP, et al. Neuroleukin: a lymphokine product of lectin-stimulated Tsells. Science 234: 574-579, 1986 Hanson PG, Flaherty DK. Immunological responses to training in conditioned runners. Clinical Science 60: 225-228, 1981 Haralambie G, Keul J. Serum glycoprotein levels in athletes in training. Experientia 26: 959-960, 1970

200

Hedfors E. Holm G. lvansen M, Wahren J. Physiological variation of blood lymphocyte reactivity: T-cell subsets, immunoglobulin production, and mixed-lymphocyte reactivity. Clinical lmmunology and lmmunopathology 27: 9-14. 1983 Hedfors E. Holm G. Ohnell B. Variations of blood lymphocytes during work studied by cell surface markers. DNA synthesis and cytotoxicity. Clinical and Experimental lmmunology 24: 328-335. 1976 Hoffman-Goetz L. Keir R, Thorne R. Houston ME. Young C. Chronic exercise stress in mice depresses splenic T lymphocyte mitogenesis i n vilro. Clinical and Experimental lmmunology 66: 551-557. 1986 Israel S. Buhl B. Krause M. Neumann G . Die Konzentration der lmmunglobuline A. G und M im Serum bci Trainierten und Untrainierten sowie nach verschiedenen sportlicken Ausdauerleistungen. Mcdiz~nund Sport 22: 225-231, 1982 Jemmott JB. Locke SE. Psychosocial factors, immunologic mcdiation. and human susceptibility to infectious diseases: how much d o we know? Psychological Bulletin 95: 78-108. 1984 Johnston RB. Monocytes and macrophages. New England Journal of Medicine 318: 747-752. 1988 Jokl E. The immunological status of athletes. Journal of Sports Medicine 14: 165-192. 1974 Keast D. Cameron K, Morton AR. Exercise and the immune response. Sports Medicine 5: 248-267. 1988 Landmann R M. Muller FB. Perini C. Wesp M. Erne P, Buhler FR. Changes in immunoregulatory cells induced by psychological and physical stress: relationship to plasma catecholamines. Clinical and Experimental lmmunology 58 : 127-1 35, 1984 Lehtonen OPJ. Tenovuo J, Aaltoncn AS, Vilja P. Immunoglobulins and innate factors of immunity in saliva of children prone to respiratory infections. Acta Pathologica Microbiologica et lmmunologica Scandinavica (Sect C) 95: 35-40, 1987 Levitt D. Cooper MD. Lymphocytes: B cells. In Stites D P et al. (Eds) Basic and clinical immunology. 6th ed., Appleton & Lange. Los Altos, CA, 1987 Lewicki R, Tchorzewski H. Denys A, Kowalska M. Golinska A. Effect of physical exercise on some parameters of immunity in conditioned sportsmen. lnternational Journal of Sports Medicine 8: 309-314. 1987 Lewicki R. Tchorzewski H, Majewska E. Nowak Z. Baj Z. Effect of maximal physical exercise on T-lymphocyte subpopulations and on interleukin I (IL I ) and interleukin 2 (IL 2) production i n vitro. lnternational Journal of Sports Medicine 9: 114-1 17, 1988 Li YS, Kouassi E, Revillard JP. Differenlial regulation of mouse B-cell activation by beta-adrenoceptor stimulation depending on type of mitogens. lmmunology 69: 367-372. 1990 Licscn H. Dufaux B, Hollman W. Modifications or' serum glycoproteins the days following a prolonged physical exercise and the influence of physical training. European Journal of Applied Physiology 37: 243-254. 1977 Liesen H. Fisher W. Dufaux B, Hollman W. Effects of endurance training on changes in serum-glycoproteins and immunoglobulins in 55 to 70 year-old men. Proceedings of the International Congress of Physical Activity Sciences. Quebec: 189. 1976 Lydyard P. Grossi C. Cells involved in the immune response. In Roitt et al. (Eds) Immunology. 2nd ed.. Gower Medical Publishing. New York, 1989 MacKinnon LT. Chick TW. van As A. Tomasi TB. The effect of exercise on secretory and natural immunity. Advances in Experimental Medicine and Biology 216A: 869-876. 1987 Madden KS. Ackerman KD. Livnat S, Felten SY, Feltcn DL. Patterns of nonadrcnergic innervation of lymphoid organs and immunological consequences of denervation. In Goetzl EJ and Spector NH (Eds) Neuroimmune networks: physiology and diseases. Alan R. Liss. Inc.. New York. 1989 Mahan MP. Young MR. Immune parameters of untrained or ex-


ercise-trained rats after exhaustive exercise. Journal of Applied Physiology 66: 282-287. 1989 Male D, Roitt 1. Adaptive and innate immunity. In Roitt 1M et al. (Eds) Immunology. 2nd ed.. Gower Medical Publishing, New York. 1989 Nehlsen-Cannarella SL. Nieman DC, Balk-Lamberton AJ. Markoff PA, Arabatzis K, et al. The effects of moderate exercise training on immune response. Medicine and Science in Sports and Exercise 23. in press. 199 la Nehlsen-Cannarella SL, Nieman DC. Jessen J. Chang L. Gusewitch G et al. The effects of acute moderate exercise on lymphocyte function and serum immunoglobulin levels. lnternational Journal of Sports Medicine 12. in press. 1991b Nieman DC, Berk LS. Simpson-Westerberg M. Arabatzis K, Youngberg WS, et al. Effects of long endurance running on immune system parameters and lymphocyte function in experienced marathoners. lnternational Journal of Sports Medicine 10: 317-323. 1989a Nieman DC. Johansen LM. Lee JW. Arabatzis K. Infectious episodes in runners before and after the Los Angeles Marathon. Medicine and Science in Sports and Exercise 20: S42. 1988 Nieman DC, Nehlsen-Cannarella SL. Markoff PA. Balk-Lamberton AJ. Yang H, et al.The effects of moderate exercise training on natural killer cells and acute upper respiratory tract infections. lnternational Journal of Sports Medicine I I, in press. 1990 Nieman DC, Tan SA. Lee JW, Berk LS. Complement and immunoglobulin levels in athletes and sedentary controls. International Journal of Sports Medicine 10: 124-128. 1989b Nossal GJV. Current conceots. immunoloev: The basic comoonents of the immune sysiem. New ~ n ~ l a nJournal d of ~ e d i cine 316: 1320-1325. 1987 Oppenheim JJ. Ruscetti FW, Faltynek CR. lnterleukins and interferons. In Stites D P et al. (Eds) Basic and clinical immunology. 6th ed., Appleton & Lange. Los Altos, CA 1987 Pahlavani MA, Cheung TH. Chesky JA. Richardson A. Influence of exercise on the immune function of rats of various ages. Journal of Applied Physiology 64: 1997-200 1. 1988 Peters EM, Bateman ED. Respiratory tract infections: an epidemiological survey. South African Medical Journal 64: 582584, 1983 Petrova IV. Kuz'min SN, Kurshakova TS. Suzdail'nitskii RS, Pershin BB. Neutrophil phagocytic activity and the humoral factors of general and local immunity under intensive physical loading. Zhurnal Mikrobiologii. Epidemiologii i Immunobiologii 12: 53-57, 1983 Poortmans JR. Serum protein determination during short exhaustive physical activity. Journal of Applied Physiology 30: 190- 192, 1970 Poortmans JR, Haralambie G. Biochemical changes in a IOOkm run: proteins in serum and urine. European Journal of Physiology 40: 245-254. 1979 Reynolds HY. Identification and role of immunoglobulins in respiratory secretions. European Journal of Respiratory Diseases 71 (Suppl.): 103-1 16. 1987 Reynolds HY. immunoglobulin G and its function in the human respiratory tract. Mayo Clinic Proceedings 63: 161-174. 1988 Ricken KH, Kindermann W. Behandlungsmoglichkeitender Infektanfalligkeit des Leistungssportiers. Deutsche Zeitschrift fur Sportsmedizin 37: 146-150, 1986 Robertson AJ. Ramesar KCRB. Potts RC. Gibbs JH. Browning MCK, et al. The effect of strenuous physical exercise on circulating blood lymphocytes and serum cortisol levels. Clinical and Laboratory lmmunology 5: 53-57, 1981 Rocker L, Kirsch KA. Stoboy H. Plasma volume, albumin and globulin concentrations and their intravascular masses: a comparative study in endurance athletes and sedentary subjects. European Journal of Applied Physiology 36: 57-64. 1976 Schouten WJ. Verschuur R, Kemper HCG. Habitual physical ac-

Exercise and lmmunoglobulins

tivity, strenuous exercise and salivary immunoglobulin A levels in young adults: the Amsterdam growth and health study. International Journal of Sports Medicine 4: 289-293, 1988 Shavit Y, Terman GW, Martin FC. Lewis JW. Liebeskind JC, et al. Stress, opioid peptides, the immune system, and cancer. Journal of lmmunology 135: 834s-837s. 1985 Sherman NA. Smith RS, Middleton E. Effect of adrenergic compounds aminophylline and hydrocortisone, on in vitro immunoglobulin synthesis by normal human peripheral lymphocytes. Journal of Allergy and Clinical lmmunology 52: 13-22, 1973 Shimokata K. Totani Y, Tsunekawa H. Sasou H. Miyachi T. Serologic study of febrile upper respiratory tract infections in adults. Japanese Journal of Medicine 27: 286-290. 1988 Stead RH, Bienenstock J. Stanisz AM. Neuropeptide regulation of mucosal immunity. lmmunological Review 100: 333-359, 1987 Stephenson LA. Kolka MA. Wilkerson JE. Effect of exercise and passive heat exposure on immunoglobulins and leucocyte concentrations. In Dotson CO & Humphrey JH (Eds) Exercise physiology: current selected research, pp. 145-157, AMS Press, Inc.. New York, 1985 Stites DP, Rodgers RPC. Clinical laboratory methods for dctection of antigens & antibodies. In Stites DP et al. (Eds) Basic and clinical immunology, 6th ed., Appleton & Lange. Los Altos, CA, 1987 Stobo JD. Lymphocytes: T cells. In Stites DP et al. (Eds) Basic and clinical immunology. 6th ed.. Appleton & Lange. Los Altos, CA, 1987 Taylor C. Rogers G. Goodman C, Baynes RD. Bothwell TH, et al. Hematologic. iron-related, and acute-phase protein responses to sustained strenuous exercise. Journal of Applied Physiology 62: 464-469. 1987

Tomasi TB, Trudeau FB, Czerwinski D, Erredge S. Immune parameters in athletes before and after strenuous exercise. Journal of Clinical lmmunology 2: 173-178, 1982 Tvede N, Heilmann C, Halkjaer-Kristensen J, Pederson BK. Mechanisms of B-lymphocyte suppression induced by acute physical exercise. Journal of Clinical and Laboratory Immunology 30: 169-173, 1989 Turner M. Molecules which recognize antigens. In Roitt IM et al. (Eds) Immunology, 2nd ed., Gower Medical Publishing. New York, 1989 von Weiss M, Fuhrmansky J, Lulay R, Weicker H. Haufigkeit und Ursache von Immunoglobulin-Mange1 bei Sportlern. Deutsche Zeitschrift fiir Sportmedizin 36: 146-153. !985 Weiss JM, Sundar SK, Becker KJ. Stress-induced immunosup pression and immunoenhancement: cellular immune changes and mechanisms. In Goetzl W & Spector NH (Eds) Neuroimmune networks: physiology and diseases, Alan R. Liss, Inc.. New York, 1989 Welliver RC, Ogra, PL. lmmunology of respiratory viral infections. Annual Review of Medicine 39: 147-162, 1988 Wells CL, Stern JR. Hecht LH. Hematologieal changes following a marathon race in male and female runners. European Journal of Applied Physiology 48: 41-49, 1982 White PAE. Ward AM. Methodological aspects of serum immunoglobulin assays. In French MAH (Ed.) lmmunoglobulins in health and disease, MTP Press Limited, Boston, 1986 Widmann FK. An introduction to clinical immunology, F.A. Davis Company. Philadelphia, 1989

Correspondence and reprints: Dr David C. Nieman. Department of Health, Leisure and Exercise Science, Appalachian State University, Boone, NC 28608, USA.