A “J”-shaped model has been proposed to describe the relationship between physical activity and risk of upper-respiratory tract infection (URTI). This model suggests that the lowest risk of URTI is found among individuals who are moderately active, and that risk of URTI is increased for both physically inactive and highly active individuals (21). It is well known that risk of URTI is increased after extreme physical exertion (i.e, running a marathon) (21) or during periods of intensive exercise training (11). However, few epidemiologic data are available to support the contention that moderately active nonathletic individuals are at lowest risk for URTI (23). To date, available evidence suggests that moderate levels of activity are associated with a reduced duration of URTI episodes (24,25), but its effect on incidence of URTI remains unclear.
Given that the average adult experiences two to five URTIs per year (12), and that these illnesses are a leading cause of physician visits and missed days of work (9), a better understanding of the relationship between physical activity and URTI is warranted. We are aware of no large-scale epidemiologic studies specifically examining the trough of the hypothesized “J”-shaped URTI–physical activity relationship in nonathletic populations, where 10% or less of the population are likely to participate in high levels of vigorous exercise training (30). Accordingly, the purpose of this investigation was to examine differences in URTI risk between inactive and moderately active individuals in a cohort of predominately nonathletic adults.
Participant recruitment and study design.
The Seasonal Variation of Blood Cholesterol Study (SEASON) was an observational longitudinal study of 641 healthy adults designed to quantify the magnitude and timing of seasonal changes in blood lipids and to identify the major factors contributing to this variation. Further details of the study design and recruitment procedures can be found in previous publications (16). Recruitment was completed between December 1994 and February 1997, and follow-up was completed in March 1998. The Institutional Review Boards of the Fallon Healthcare System and the University of Massachusetts Medical School approved all participant recruitment and data collection procedures. Each participant read and signed an approved informed consent. At baseline and in each of four subsequent quarters of follow-up (at 90-d intervals), individuals came to the clinic to provide blood samples, have their body mass measured, and to return a series of self-administered questionnaires. Physical activity, diet, and light exposure data were collected using three 24-h recall (24HR) telephone interviews at each assessment point. 24HR interviews were conducted within a 42-d call window (−28 to +14 d) surrounding each clinic visit.
Demographic data (e.g., age, gender, marital status, education, and employment), the number and ages of household members, health habits (smoking history, vitamin C, and multivitamin supplementation (13), and minutes of exercise each week (19)) were collected by self-administered questionnaires at study baseline. Anthropometric data, including body mass (kg) and height (m), were measured during clinic each visit, and body mass index (BMI) was calculated (kg·m−2). Classifications by BMI status (i.e., normal weight, overweight, and obese) were made using recently published guidelines for the Detection and Treatment of Obesity (20).
Information regarding URTI was collected by interview at each clinic visit. Participants were asked to report the number of colds, flu, or allergic episodes they experienced in the 3 months preceding their clinic visit using their common understanding of these terms. Explicit symptoms (e.g., runny nose, and cough) and the duration (in days) of the events were not routinely recorded. Reports of more than one episode of illness in a 3-month period of recall were possible. The reported number of colds was used to estimate incidence of URTI in our population. This methodology was comparable to the other studies of physical activity and URTI that also relied on self-report of past illness. Previous studies have utilized recall periods of ranging from 1 d (daily logs) to up to 6 months (22). Although we did not have explicit reports of URTI symptoms in our assessment, the ability to report allergy and flu events separately allowed subjects to differentiate colds from episodes of allergy and from more severe flu-like illnesses.
Physical activity assessment.
Physical activity was assessed by a series of 15 24-h physical activity recalls (24PAR). The 24PAR assessment, as well as relative validity studies of the method, have been described in detail elsewhere (18). Briefly, trained registered dieticians conducted unannounced telephone-administered interviews on two randomly selected weekdays and one randomly selected weekend day during the call window surrounding each clinic visit. In the interview, participants recalled the number of hours they spent in four intensities of activity in the previous day (light, moderate, vigorous, and very vigorous), in each of three activity domains (household, occupational, and leisure time). Standard methods were employed to calculate estimates of physical activity energy expenditure (metabolic equivalent (MET)-h·d−1) using standard MET values and reported duration (h·d−1) of activity (18). Summary scores using the average of all 24PAR were calculated after weighting weekday and weekend day for their sampling frequency. For these analyses only activities of 3 METs or more were examined (i.e., moderate-vigorous activity), and moderate (3.0–5.9 METs) and vigorous (6+ METs) were evaluated separately. Completion rates of 24PAR recalls were high (mean number of recalls per subject over the 12-month study period = 12.3, SD = 2.8).
Three 24-h dietary recalls were also administered in each quarter of follow-up. The 24-h dietary recall data were collected using Nutrition Data System data entry and nutrient database software developed and maintained by the Nutrition Coordinating Center at the University of Minnesota, Minneapolis, MN (8). Dietary recalls of less than 750 kcal·d−1 or more than 5000 kcal·d−1, or judged by the interviewer to be unreliably reported were excluded from the present analyses (<4% of diet recalls). Dietary variables considered in these analyses were total energy intake (kcal·d−1), the energy contributing nutrients [carbohydrate, protein, total fat, and alcohol (g·d−1, %energy)], and selected dietary constituents that have been associated with lower URTI risk (i.e., vitamin C and glutamic acid) (6,14). The 1-yr average dietary intakes (foods and supplements) were employed as continuous variables in the analyses.
At baseline and in each quarter of follow-up, participants completed self-administered versions of the Beck Depression (4,5) and Anxiety (3) inventories, as well as a subscale of the Cook-Medley Hostility Scales that focuses on cynicism (2). One-year average scores derived from these scales were employed to control for potential confounding by psychosocial factors.
Descriptive characteristics of the cohort were evaluated in men and women. Physical activity patterns across quartiles of moderate-vigorous activity were also examined. Descriptive results were tested for differences by gender and activity levels by using analysis of variance and chi-square methods. Gender-specific quartile cut-points for 1-yr average moderate-vigorous activity in men were 3.93, 7.15, and 11.95 MET-h·d−1, and in women were 2.38, 4.09, and 6.24 MET-h·d−1, respectively. The number of colds reported and months at risk for URTI were calculated over the full year of follow-up (annually) and in each season. URTI assessment dates were classified into seasonal categories using common seasonal cut-points (i.e., winter = December 21 to March 20, spring = March 21 to June 20, summer = June 21 to September 20, and fall = September 21 to December 20).
Poisson maximum likelihood regression was used to estimate incidence rate ratios for URTI across gender-specific quartiles of physical activity while normalizing URTI rates for the number of months at risk in the period of interest (29). The dependent variable in these models was the number of colds reported in the period of interest. Independent variables, moderate-vigorous physical activity quartiles, were fit as dummy coded variables, with the least active quartile serving as the referent category. Tests for linear trend in the rates across quartiles were completed by fitting the activity quartiles as a continuous variable (i.e., values from 1 to 4). Covariates were fit as continuous or dummy coded categorical variables. Potential confounders considered were age, BMI, demographic variables (education, family number, family structure [e.g., young children]), smoking, vitamin supplementation (vitamin C and multivitamins), dietary intake (macronutrients, alcohol, vitamin C, and glutamic acid), and the psychosocial variables (anxiety, cynicism, and depression). Models adjusting only for age and fully adjusted models (covariates with P < 0.10) were fit to estimate the effect of physical activity on URTI risk in all subjects, men and women separately, and in each season of the year. Exclusions from the present analyses were made for participants with no 24PAR data (N = 5) for having fewer than two quarters of study participation (N = 56) or having missing URTI information (N = 33).
Descriptive characteristics of the SEASON cohort are presented in Table 1. Men and women reported 1.2 (1.4) and 1.2 (1.2) URTI events per year, respectively [mean (SD)]. Figure 1 describes seasonal variation of the reported incidence of colds (URTI), flu, and allergy in the SEASON cohort. Approximately 40% of the cohort reported URTI events in the winter, 10% in the summer, and 40% again by late fall (Fig. 1).
Table 2 describes the patterns of physical activity within each quartile of moderate-vigorous activity. Moderate-intensity activity (3.0–5.9 METs) constituted the primary activity intensity category that increased overall activity levels across quartiles in both men and women. There were no differences in light activities across quartiles of moderate-vigorous activity in either men or women. Among men, the occupational and household domains contributed most to increasing moderate-vigorous activity levels. Among women, there was a generalized increase in each of household, occupational, and leisure-time activity domains. Men and women in the two upper quartiles reported an average of about 60–80 min·wk−1 in overt exercise activities. Only 19 men (14%) and 24 women (19%) in the two upper moderate-vigorous activity quartiles reported participating in more than 150 min·wk−1 of exercise (data not shown).
Among all participants, a 29% reduction in URTI risk was observed in comparing the upper to the lower quartile of activity, after controlling for age and gender (Table 3). Further adjustment of this model for education, anxiety, cynicism, and %fat intake attenuated the estimate only slightly (Table 3). The effect of moderate-vigorous activity on URTI risk appeared to be stronger in men than women, and with the reduction of sample size in gender-specific analyses, was only statistically significant among men. In fully adjusted models within each season of the year, the majority of the reduction of risk in the annual URTI incidence appeared to be realized in the summer and fall months (Fig. 2).
In detailed analyses, the effects of other covariates (BMI, number of adults in household, presence of children in the household, smoking, vitamin supplementation, macro- and micro-nutrient intake, and depression) were not associated with URTI risk, nor did they modify the observed relationship between activity and URTI risk. We also examined moderate- and vigorous-intensity activities, and each of the activity domains (household, occupational, and leisure) separately, and did not observe the consistent risk reductions noted for overall moderate-vigorous activity in the analyses presented in Table 3. In addition, we did not observe an increased risk of URTI with either high levels of vigorous activity or high levels of leisure-time activity in this middle-aged cohort of predominantly nonathletic adults.
Findings from the present investigation demonstrate that high levels of moderate-vigorous activity accumulated at home, at work, and during leisure-time were associated with a 20–30% reduction in the annual risk of URTI in a cohort of predominantly nonathletic adults. Reductions in URTI risk were particularly evident in the summer and fall seasons. These data support the hypothesis that there is a “J”-shaped relationship between physical activity and infection risk. Specifically, we observed moderate levels of physical activity to be associated with a reduced risk for URTI relative to low levels of activity. Because relatively few individuals in this cohort were participating in high levels of intense exercise training, we were unable to sufficiently evaluate the effects of high levels of exercise training on URTI risk.
It is clear from observational and laboratory studies that high levels of exercise training and athletic competition are associated with increased risk of URTI. For example, Peters and Bateman (27) were the first to report that marathon participants had a higher incidence of URTI after running a marathon in comparison with nonracing controls (33% vs 15%, respectively). Since that publication, at least 10 studies have examined the effect of exercise training on URTIs, primarily among competitive athletes (runners). The majority of these studies have found more elite athletes to have higher incidence of URTI, and this has been most consistently associated with higher levels of training (12,22).
In contrast, previous studies supporting the contention that moderate levels of physical activity are associated with reduced risk of URTI incidence have been mixed. Neiman and colleagues (24–26), using randomized designs, studied the effects of walking 4–5 times per week for 30–45 min·d−1. Among older women (65–85 yr), regular walking was associated with a markedly reduced, but statistically nonsignificant, incidence of self-reported URTI (21% vs 50%, respectively). In an age-matched highly active but nonrandomized comparison group, the incidence of URTI was 8% during the same period (24). Among younger women (25–45 yr), 15 wk of regular walking resulted in no difference in URTI incidence in comparison with a control group, but the walkers reported half as many days of URTI symptoms during the study period (3.7 vs 7.0 d, P = 0.05) (26). In the only large epidemiologic study published examining the relationship between URTI and physical activity (N = 176) of which we are aware, Schouten and colleagues (28) reported weak inverse relationships between total sports activity and self-reported URTI incidence in young men and women (Spearman r = −0.12 and −0.18, respectively). This finding was statistically significant only in women. The present investigation extends this body of literature by demonstrating that moderate to vigorous physical activity was associated with a 20–30% reduction in URTI risk after adjusting for many of the known risk factors for infection.
Physical activity has been noted to beneficially modify immune function in a manner that would be consistent with our findings. Acute moderate intensity activity temporarily enhances a number of immune parameters (neutrophils, circulating lymphocyte number, natural killer cells, and cytokines), and these effects may last for several hours after the bout of physical activity (15). Additionally, chronic exercise training has been noted to increase resting natural killer cell activity (15). These modulations in immune function may be responsible for the observed reductions in URTI in risk in the present investigation. Our lack of finding of an increased risk of URTI in the most active groups in our cohort likely stems from the fact that the characteristics of moderate-vigorous household and occupational activity examined in the present study differ considerably from high-level exercise training or physical exertion (i.e., training for or running a marathon) that typifies activities associated with immune suppression and increased URTI risk (22). Household and occupational activities in which this group of middle-aged adults participated were probably done at moderate intensities (50–70% V̇O2 max), over a number of hours each day, and on many days of the week. This type of activity pattern would be predicted to benefit from the repeated temporary effects on the immune system that have been established for acute physical activity and, perhaps also, maximize the known chronic effects of activity on the immune system (15).
A limitation of the present investigation was that it was an observational study that relied on self-reports of both the URTI outcome measure and the exposure of main interest (physical activity). Although the validity of the instrument we employed to capture URTI events has not been formally tested, our estimates of the annual and seasonal variation in URTI rates were virtually identical to previously published reports that utilized more intensive assessment methods. For example, Heath and colleagues (11) reported an annual incidence of URTI of 1.2 events per year in a group of runners who kept daily logs of their URTI symptoms for a full year. Gwaltney and colleagues (10) reported, again using daily records of URTI symptoms, that seasonal variation in URTI fluctuated from 30 to 40% in the winter to roughly 10% in the summer months. The annual rates of URTI were reported to be stable between 1982 and 1996 (1), suggesting little secular variation in these outcomes. Overall, these findings lend support to the validity of our URTI assessment.
The utility of the 24PAR has been established in a validation study by comparison with objective and other self-report instruments (18), as well as in an investigation of the seasonal variation in physical activity patterns in this cohort (16). Our examination of the sources of variance in daily physical activity in this cohort has demonstrated that roughly 85% of the interindividual difference in total activity in men would be captured by 14 d of 24PAR, whereas only about 75% of the true differences in activity in women would be accounted for by 14 d of 24PAR (17). Thus, the 1-yr average of multiple 24PAR employed in these analyses appears to have been useful in reliably separating physically inactive individuals from the most active individuals. To the extent that our measure did not reliably separate individuals, our risk estimates would be predicted to be biased toward the null because of the presence of random variation in our activity measures. Indeed, our previous finding of a lower reliability in the 24PAR for total activity in women (17) suggests that the weaker URTI–activity association observed in women could be attributed to measurement error.
URTIs are primarily caused by viruses that produce relatively mild symptoms (e.g., runny nose, cough, and sore throat) and are routinely referred to as “common colds” and flu (7,12). The incidence of UTRI is quite high among adults (1-5 illnesses per year) (7,9,11), and these infections are the major contributor to morbidity in Western countries. For example, Couch (7) estimated that 100 million colds per year in the United States were responsible for 250 million days of restricted activity and 30 million days of lost work (7). Clearly, the effect of URTIs on health care cost and worker productivity is substantial. The present findings support the hypothesis that moderate levels of physical activity can reduce the risk of URTI by 20–30%. These data provide additional weight to the growing body of evidence that participation in regular physical activity can reduce both morbidity and mortality at the population level (30).
The authors would like to thank Laura Robidoux for her assistance with study recruitment and the clinic-based data collection; Donna Gallagher for her assistance with study mailings; Kelly Scribner for coordination of the 24-h recalls; and the SEASON dieticians who conducted the recalls. Additionally, the authors would like to thank Yunsheng Ma and Thomas Hurley for their organizational and data management expertise.
This work was supported by the American College of Sports Medicine’s Fellowship for Epidemiological Research on Physical Activity and Health (1998) to Charles E. Matthews, Ph.D., and the National Heart, Lung, and Blood Institute, grant no. RO1 HL-52745 to Ira S. Ockene, M.D.
Address for correspondence: Charles E. Matthews, Ph.D., Department of Epidemiology and Biostatistics, University of South Carolina, Norman J. Arnold School of Public Health, Columbia, SC 29208; E-mail: [email protected]
1. Adams, P. F., G. E. Hendershot, and M. A. Marano. Current estimates from the National Health Interview Survey, 1996. Vital Health Statist. 10: 1–10, 1996.
2. Barefoot, J. C., K. A. Dodge, B. L. Peterson, W. G. Dahlstrom, and R. B. Williams, Jr. The Cook-Medley hostility scale: item content and ability to predict survival. Psychosom. Med. 51: 46–57, 1989.
3. Beck, A. T., N. Epstein, G. Brown, and R. A. Steer. An inventory for measuring clinical anxiety: psychometric properties. J. Consult. Clin. Psychol. 56: 893–897, 1988.
4. Beck, A. T., C. H. Ward, M. Mendelson, J. Mock, and J. Erbaugh. An inventory for measuring depression. Arch. Gen. Psychiatry 4: 561–570, 1961.
5. Beckham, E. E., and W. R. Leber, (Eds.). Handbook of Depression: Treatment, Assessment, and Research. Homewood, IL: Dorsey Press, 1985, pp. 985–986.
6. Castell, L. M., and E. A. Newsholme. Glutamine and the effects of exhaustive exercise upon the immune response. Can. J. Physiol. Pharmacol. 76: 524–532, 1998.
7. Couch, R. The common cold: control? J. Infect. Dis. 150: 167–173, 1984.
8. Feskanich, D., B. H. Sielaff, K. Chong, and I. M. Buzzard. Computerized collection and analysis of dietary intake information. Comput. Methods Programs Biomed. 30: 47–57, 1989.
9. Grahm, N. The epidemiology of acute respiratory infections in children and adults: a global perspective. Epidemiol. Rev. 12: 149–178, 1990.
10. Gwaltney, J., J. Hendley, G. Simon, and W. Jordan. Rhinovirus infections in an industrial population: I. The occurrence of illness. N. Engl. J. Med. 275: 1261–1268, 1966.
11. Heath, G., E. Ford, T. Craven, C. Macera, K. Jackson, and R. Pate. Exercise and the incidence of upper respiratory tract infections. Med. Sci. Sports Exerc. 23: 152–157, 1991.
12. Heath, G. W., C. A. Macera, and D. C. Nieman. Exercise and upper respiratory tract infections: is there a relationship? Sports Med. 14: 353–365, 1992.
13. Hebert, J., I. Ockene, T. Hurley, R. Luippold, A. Well, and M. Harmatz. Development and testing of a seven-day dietary recall. J. Clin. Epidemiol. 50: 925–937, 1997.
14. Hemila, H. Vitamin C and common cold incidence: a review of studies with subjects under heavy physical stress. Int. J. Sports Med. 17: 379–383, 1996.
15. Mackinnon, L. T. Advances in Exercise Immunology. Champaign, IL: Human Kinetics, 1999, pp. 55–112.
16. Matthews, C. E., P. S. Freedson, I. E. J. Stanek, et al. Seasonal variation of household, occupational, and leisure-time physical activity: longitudinal analyses from the Seasonal Variation of Blood Cholesterol Study. Am. J. Epidemiol. 153: 172–183, 2001.
17. Matthews, C. E., J. R. Hebert, P. S. Freedson, et al. Sources of variance in daily physical activity levels in the Seasonal Variation of Blood Cholesterol Study. Am. J. Epidemiol. 153: 987–995, 2001.
18. Matthews, C. E., J. R. Hebert, P. S. Freedson, E. J. Stanek, I. S. Ockene, and P. A. Merriam. Comparing physical activity assessment methods in the Seasonal Variation of Blood Cholesterol Levels Study. Med. Sci. Sports Exerc. 32: 976–984, 2000.
19. Matthews, C. E., J. R. Hebert, I. S. Ockene, G. Saperia, and P. A. Merriam. Relationship between leisure-time physical activity and selected dietary variables in the Worcester Area Trial for Counseling in Hyperlipidemia. Med. Sci. Sports Exerc. 29: 1199–1207, 1997.
20. N.H.L.B.I. Clinical Guidelines on the Identification, Evaluation, and Treatment of Overweight and Obesity IN Adults: The Evidence Report. Bethesda, MD: National Institutes of Health; National Heart, Lung, and Blood Institute, 1998, p. 60.
21. Nieman, D. C. Exercise, infection, and immunity. Int. J. Sports Med. 15: S131–141, 1994.
22. Nieman, D. C. Exercise, upper respiratory tract infection, and the immune system. Med. Sci. Sports Exerc. 26: 128–139, 1994.
23. Nieman, D. C. Is infection risk related to exercise workload? Med. Sci. Sports Exerc. 32: S406–S411, 2000.
24. Nieman, D. C., D. A. Henson, G. Gusewitch, et al. Physical activity and immune function in elderly women. Med. Sci. Sports Exerc. 25: 823–831, 1993.
25. Nieman, D. C., S. L. Nehlsen-Cannarella, D. A. Henson, et al. Immune response to exercise training and/or energy restriction in obese women. Med. Sci. Sports Exerc. 30: 679–686, 1998.
26. Nieman, D. C., S. L. Nehlsen-Cannarella, P. A. Markoff, et al. The effects of moderate exercise training on natural killer cells and acute upper respiratory tract infections. Int. J. Sports Med. 11: 467–473, 1990.
27. Peters, E. M., and E. D. Bateman. Ultramarathon running and upper respiratory tract infections: an epidemiological survey. South Afr. Med. J. 64: 582–584, 1983.
28. Shouten, W., R. Verschur, and H. Kemper. Physical activity and upper respiratory infections in a normal population of young men and women: the Amsterdam Growth and Health Study. Int. J. Sports Med. 9: 451–455, 1988.
29. Stata. Stata Statistical Software: Release 4.0. College Station, TX: Stata Corporation, 1995, pp. 69–77.
30. U.S.D.H.H.S. Physical Activity and Health: A Report of the Surgeon General. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Publication S/N 017-023-00196-5, 1996, p. 4.