High levels of air pollution in Beijing leading up to the 2008 Olympic Games (22,35) raised questions about possible effects on athletes' health and performance (33). Currently, there are no known documented measures of air pollutant levels during major marathons. On the basis of scientific studies documenting air pollutants' effects on health, the US Environmental Protection Agency (EPA) has established the National Ambient Air Quality Standards (NAAQS) for criteria pollutants, listed in Table 1. Air quality in most major metropolitan areas typically exceeds the standards at least once per year. Fortunately, for runners, most marathons are held on weekend mornings, when vehicular activity and its associated emissions are low, and photochemical reactions driven by solar radiation have not yet produced secondary pollutants such as ozone (O3).
;)
TABLE 1: Pollutant concentrations during seven major marathons, approximate US urban background, EPA standards, and minimum levels known to affect lung function during exercise.
Laboratory studies have identified thresholds of singular pollutants that are associated with lung function decrements during exercise (9,18,29). For O3, the threshold is 2.4 times lower than the 1-h standard, whereas for nitrogen dioxide (NO2) and sulfur dioxide (SO2), the thresholds are 19 and 1.4 times higher than their respective standards (Table 1). The laboratory studies used either low exercise intensities (≤65% V˙O2max) (11,12,23,28,36) or short exposure durations (∼1 h) (11,12,29). The resulting information indicates an inverse dose-response relationship between air pollutants and lung function at constant workloads (11,12,28,29,32,36) as well as an inverse dose-response relationship between lung function and breathing frequency at constant pollutant levels above previously identified thresholds (1,5,12).
It is possible that marathon runners will be affected by lower pollutant concentrations than those in the laboratory studies. Marathon foot races (42 km) represent situations in which there are short-term exposures to air pollutants under atypical breathing conditions. An athlete running at 70% of maximal oxygen uptake for the length of a marathon (∼3 h) inhales the same volume of air as a sedentary person would in 2 d (4). In addition to the elevated ventilation rate (7,32), the switch from nasal to mouth breathing (7,25) and an increased airflow velocity carry pollutants deeper into the lungs (24) and further amplify the runner's dose of pollutants.
Studies have also suggested that females may be more susceptible to the negative effects of air pollution than males, which could be due to a higher fractional deposition of pollutant particles in the airways (6,27). In a study of beach lifeguards in Texas, lung function decrements were significantly larger for women than men with increased levels of ambient O3 and fine particulate matter (PM2.5) (36). Similarly, Shima and Adachi (34) noted that females had increased airway hyperresponsiveness to NO2 than males. This is not a universal finding because others have found no differences in lung function between sexes when exercising at similar relative workloads and pollutant concentrations (21).
The purposes of this study were to 1) describe the concentrations of carbon monoxide (CO), O3, particulate matter of 10 μm and less (PM10), particulate matter of 2.5 μm and less (PM2.5), NO2, and SO2 present during major US marathons; 2) quantify performance decrements associated with air pollutants; and 3) examine any sex difference in performance related to air pollutants. The selected pollutants comprise the US EPA's list of "criteria pollutants" (minus lead), whose health and environmental effects are well established. It was hypothesized that the low levels of air pollution during marathons did not affect performance for either men or women.
METHODS
Marathon racing performances for the top three men and women were collected from seven competitive US races: Boston (Boston, MA), New York (New York, NY), Chicago (Chicago, IL), Twin Cities (Minneapolis/St. Paul, MN), Grandma's (Duluth, MN), California International (Sacramento, CA), and Los Angeles (Los Angeles, CA) marathons. These data are in the public domain; therefore, written and informed consent was not required from individual athletes. Finishing data up through the year 2007 were available for 28, 28, 28, 26, 25, 25, and 8 consecutive years, respectively. Performance was assessed by calculating a finishing time as a percent off course record for the average time of the first three male and female finishers [(finishing time − course record)/course record × 100]. The course record used was always the current course record for the year under study. This approach has been used previously to analyze performance (15).
Air pollutant concentrations corresponding to the race day, time, and location of the marathon were collected from the EPA's online Air Quality System database. Pollutants were measured according to EPA's Federal Reference Methods: infrared absorption for CO, ultraviolet absorption for O3; gravimetric filters, beta attenuation monitoring, and tapered element oscillating microbalances for PM10; gravimetric filters for PM2.5; chemiluminescence for NO2; and ultraviolet fluorescence for SO2. Gaseous pollutant concentrations (CO, O3, NO2, and SO2) represent 2-h averages during the race, and particulate concentrations (PM10 and PM2.5) represent 24-h averages on the day of the race. Hourly weather data were obtained through the Air Force Combat Climatology Center. The weather data examined included dry bulb (Td), wet bulb (Tw), and black globe (Tg) temperatures. The wet bulb globe temperature (WBGT) was calculated from the Td, Tw, and Tg (38) and was averaged for the duration of each race.
Statistical analysis.
One purpose of this study was to describe the ambient air pollution concentrations during US marathons. Simple descriptive statistics, namely, mean, SD, and range, were used to illustrate the pollutant levels during the seven marathons. A second purpose of this study was to quantify any decrement of marathon performance associated with ambient air pollution. Initially, a correlation analysis was performed between the performances of men and women marathon runners, WBGT, and the six air pollutants. Because the relationship between marathon performance decrements and WBGT has been well established (15) and because this article uses much of the same data (4 of the 7 marathons or 107 of the 168 race years), it was possible to examine the effects of pollutants on performance independent of WBGT. A WBGT-adjusted performance decrement was calculated as the residual of a quadratic fit of performance versus WBGT. Least-squares linear regression analysis was then used to quantify additional performance decrements associated with air pollutants. All analyses were performed using the software program JMP (SAS Institute, Inc., Cary, NC) for curve fitting. Results were deemed significant if the P < 0.05.
RESULTS
There were 168 race years analyzed in the present data set. Four race years were excluded because of heavy wind and rain, and one race year was excluded because of a course change. Women's marathon performance ranged from −6.1% (where the top three finishers' average time was faster than the course record) to 13.9% slower than course record, and men's performance ranged from −2.1% to 8.7%. WBGT ranged from 0.1°C to 28°C for all marathons examined. Concurrent measurements of all six "criteria" pollutants (CO, O3, PM10. PM2.5, NO2, and SO2) were only available for eight race years because air pollution monitoring sites typically measure only a subset of pollutants and may not have been operational in all years. In addition, there was limited availability of PM10 because protocol calls for measurements only once every 6 d and of PM2.5 because it was not required before 1997. Therefore, data were available for CO, O3, PM10, PM2.5, NO2, and SO2 for 139, 122, 37, 41, 97, and 108 race years, respectively.
The air pollutant concentrations of CO, O3, PM10, PM2.5, NO2, and SO2 ranged from 0 to 5.9 ppm, from 0 to 0.07 ppm, from 4.5 to 41.0 μg·m−3, from 2.8 to 42.0 μg·m−3, from 0 to 0.06 ppm, and from 0 to 0.05 ppm, respectively. In no time did CO, PM10, O3, or SO2 exceed the current NAAQS (Table 1). There were two times that NO2 (Boston 1980; New York 1993) and five times that PM2.5 (California International 2000, 2002, 2004, 2006; Los Angeles 2002) exceeded the NAAQS. The lowest levels of air pollutants known to affect lung function during exercise have been reported for O3 (9), NO2 (17), and SO2 (29) (Table 1). Only five of the races experienced O3 concentrations above the threshold of 0.05 ppm (California International 1988; Boston 1985, 1993; New York 1990; Grandma's 2003), and in no races did air quality exceed 1.0 ppm for NO2 or 0.2 ppm for SO2 (Table 1).
A correlation matrix between women's and men's marathon performance, pollutant concentrations, and WBGT is presented in Table 2. There are significant correlations between men's and women's performance and WBGT and PM10. The correlations indicate that as WBGT and PM10 increase, the marathon performances of men and women become slower. As the progressively slowing of marathon performance with increasing WBGT has been established (15), an adjusted performance decrement was calculated by regression analysis using a quadratic fit (Table 3). The optimal racing WBGT is equal to 12.6°C for women and 12.7°C for men (previously reported in Ely et al. 14). The regression predicts that WBGT accounts for 6% of the variation in running performance for women and 12% for men. The remaining variations (residuals) are plotted against pollutant concentrations for women (Fig. 1) and men (Fig. 2). Further regression analysis between the residuals and pollutant concentrations (Table 4) revealed a significant correlation between PM10 and the variability in women's performance (Fig. 1 and Table 4) where for every 10-μg·m−3 increase in PM10, performance can be expected to decrease 1.4%. Men's and women's performances were not significantly correlated with any other pollutant.
TABLE 2: Correlation matrix and significance level (*<0.05) between marathon performance (percent off the course record) for women and men with air pollutants and WBGT.
TABLE 3: Quadratic regression statistics and significance of coefficients (*<0.05) for marathon performance (percent off the course record) versus WBGT (°C).
TABLE 4: Linear regression statistics and significance (*<0.05) for marathon performance (WBGT-adjusted percent off the course record) versus air pollutant concentrations.
FIGURE 1: Women's WBGT-adjusted performance (percent off course record) versus concentrations of CO, O3, PM10, PM2.5, NO2, and SO2 during the race. The least-squares linear regression line is shown for PM10. Dotted lines indicate the NAAQS of each pollutant; the x-axes for O3, PM10, and SO2 have a broken scale.
FIGURE 2: Men's WBGT-adjusted performance (percent off course record) versus concentrations of CO, O3, PM10, PM2.5, NO2, and SO2 during the race. Dotted lines indicate the NAAQS of each pollutant; the x-axes for O3, PM10, and SO2 have a broken scale.
DISCUSSION
The main finding of this investigation was that the level of pollutants present during the 168 race years analyzed rarely exceeded the health limits set by the EPA or the levels known to affect lung function during exercise in laboratory situations. Regardless, the low levels of PM10 were significantly associated with decreased performance for women but not for men. For every 10-μg·m−3 increase in PM10, women's marathon performance can be expected to decrease 1.4%. Men's and women's performances were not associated with any other pollutant.
The lack of a significant relationship between CO, O3, NO2, and SO2 and marathon performance is not surprising because the concentrations considered here were much lower than the levels that affect lung function in laboratory studies (as reviewed by Atkinson [4], Carlisle and Sharp [10], and Florida-James et al. [16]). Pollutant concentrations are determined by emissions, chemical reactions in the atmosphere, and meteorological conditions, including atmospheric mixing height and solar radiation. Vehicles are the dominant source of most pollutants in urban areas, and intense solar radiation is needed to produce O3. On weekdays, most pollutant concentrations peak in the morning because of rush hour traffic and a low mixing height through which to dilute pollutants. Except for Boston, all marathons in the present study occurred in the morning hours on weekend days, when pollutant concentrations tend to be lower because of minimal traffic and low solar radiation. The Boston Marathon, which has a midmorning start time in the early spring, in addition to the city's high latitude, minimized the potential for O3 formation. In sum, pollutant levels during marathons would not be expected to be elevated much above the urban background.
Only two known field studies have assessed performance and air pollution. Wayne et al. (37) documented a strong relationship (r = 0.88) between increasing O3 (up to 0.3 ppm) and a moderate relationship (r = 0.62) between total particle loading, an older measurement on the basis of optical reflectance of collected particles, with high school cross-country runners who failed to show an expected improvement in running time from one meet to the next over the season. In addition, a longitudinal study of thoroughbred horse race performances spanning 35 years documented that race times were markedly slower when run under O3 conditions described as hazardous (O3 > 0.405 ppm), with no changes in race performance with increasing particulate matter (0-500 μg·m−3) (19). The positive relationships between air pollutant concentration and performance in these studies are most likely due to the higher pollutant concentrations than those measured in the present data set.
In the present study, there was a significant relationship between PM10 and the performance of women (r2 = 0.33, P < 0.05; Fig. 1) but not men. The sex difference may not be a surprising finding because others have found similar trends (36). It has been suggested that females have increased airway hyperresponsiveness to air pollutants than males (34) possibly because of higher fractional deposition of pollutant particles in their airways (6,27). Women have smaller openings of the larynx (13), which could increase the turbulence of inhaled air, in addition to ∼32% smaller cross-sectional area of the trachea (31), both of which will enhance deposition of coarse particles by inertial impaction in the upper airways. The work of Alarie (2) indicates that inhaled pollutants stimulate the vagal reflex, which reduces the depth of inhalation at the expense of increasing breathing frequency at constant workloads. The increased breathing frequency, in addition to a higher fractional deposition of pollutant particles in the airways of females compared with males (6,27), may have increased subjective discomfort (1,9,20) enough for women to slow their pace.
A point of interest was that although PM2.5 is a subset of PM10, i.e., measurements of particles smaller than 10 μm encompass those smaller than 2.5 μm, only PM10 was found to be correlated with performance. Size-dependent differences in particle deposition along the respiratory tract could partially explain the result. Fluid flow models show that PM10 deposition efficiency in the nasal cavity is two to six times greater than that of PM2.5 and that the discrepancy increases with ventilation rates up to 60 L·min−1 (26,30). The higher deposition indicates that, under normal breathing conditions, a larger fraction of inhaled PM2.5 is in contact with deeper airways and the lungs on a regular basis while PM10 is filtered. During running, however, mouth breathing circumvents removal of PM10 and therefore amplifies the dose of PM10.
One potential limitation of this study is that the temperatures present during marathon running were warm but never hot. A study by Folinsbee et al. (17) documented that constant levels of air pollution caused incremental decrements in lung function with increasing heat stress (WBGT = 18°C, 26°C, 29°C, 33°C). Folinsbee et al. (17) used a range of WBGT that exceeded those found in the present study (WBGT = 28°C). Another possible limitation is that marathon running may not be ideal for the measurement of performance decrements due to air pollutants because many marathons are held in the time of year when ambient temperatures are low and solar radiation is limited, both of which are associated with low pollutant levels. A third limitation is that the study did not consider multipollutant interactions due to the six pollutant concentrations being unavailable for all marathons.
Although the 2008 Beijing Olympic marathon was not part of the current study, concern over athlete's performance in the race is what spurred this research. It is interesting to note that despite relatively high PM10 concentrations of 87 μg·m−3 on race day (3,8), the men's marathon winner set a new Olympic record. In addition, the average of the top three men's finishing times was faster than the preexisting record. During the women's marathon, PM10 concentrations averaged 62 μg·m−3 (3,8), and the top three women were 2.6% slower than the Olympics marathon record.
CONCLUSIONS
The levels of six EPA "criteria" air pollutants present during 168 race years of US marathons rarely exceeded the health limits set forth by the EPA or the levels known to affect lung function in laboratory situations. The levels of CO, O3, PM2.5, NO2, and SO2 did not seem to affect the marathon performance of men or women. Women's marathon performances were negatively affected by coarse particulate matter (PM10); for every 10-μg·m−3 increase in PM10 concentration, performance slowed 1.4%.
This work was supported by the National Science Foundation's Advance program at Virginia Tech. The authors thank TSgt Michael Ross (AFCCC), Jack Fleming (Boston Athletic Association), Shane Bauer (Grandma's Marathon), Will Nicholson (MS4 University of Minnesota Medical School), Myles Killar (Virginia Tech), and Nick Mangus (US EPA) for their assistance.
The opinions or assertions contained herein are the private views of the authors and are not to be construed as official or reflecting the views of the Army or the Department of Defense. Any citations of commercial organizations and trade names in this report do not constitute an official Department of the Army endorsement or approval of the products or services of these organizations. The authors do not have any relationships to disclose that would cause a conflict of interests. The results of the present study do not constitute endorsement by the American College of Sports Medicine.
REFERENCES
1. Adams WC, Schelegle ES. Ozone and high ventilation effects on pulmonary function and endurance performance.
J Appl Physiol. 1983;55:805-12.
2. Alarie Y. Sensory irritation by airborne chemicals.
Crit Rev Toxicol. 1973;2:299-363.
3. Associated Press. Sensing Air Quality at the Olympics. 2008 [cited 2008 Nov 7]. Available from:
http://hosted.ap.org/specials/interactives/_international/oly_fea_pollution/index.html.
4. Atkinson G. Air pollution and exercise.
Sports Exerc Inj. 1997;3:2-8.
5. Bates DV, Bell G, Burnham C, Hazucha MJ, Pengelly LD, Silverman FS. Short tern effects of ozone on the lung.
J Appl Physiol. 1972;32:176-81.
6. Bennett WD, Zeman KL, Kim C. Variability of fine particle deposition in healthy adults: effect of age and gender.
Am J Respir Crit Care Med. 1996;153:1641-7.
7. Bowes SM, Francis M, Laube BL, Frank R. Acute exposure to acid fog: influence of breathing pattern on effective dose.
Am Ind Hyg Assoc J. 1995;56:143-50.
8. British Broadcasting Corporation. In pictures: Beijing pollution-watch. 2008 [cited 2008 Nov 7]. Available from:
http://news.bbc.co.uk/2/hi/in_pictures/7506925.stm.
9. Brunekreef B, Hoek G, Breugelmans O, Leentvaar M. Respiratory effects of low-level photochemical air pollution in amateur cyclists.
Am J Respir Crit Care Med. 1994;150:962-6.
10. Carlisle AJ, Sharp NCC. Exercise and ambient air pollution.
Br J Sport Med. 2001;35:214-22.
11. Castillejos M, Gold DR, Damokosh AI, et al. Acute effects of ozone on the pulmonary function of exercising school children from Mexico City.
Am J Respir Crit Care Med. 1995;152:1501-7.
12. Delucia AJ, Adams WC. The effects of O
3 inhalation during exercise on pulmonary function and blood chemistry.
J Appl Physiol. 1977;43:75-81.
13. Eckel HE, Sittel C, Zorowka P, Jerke A. Dimensions of the laryngeal framework in adults.
Surg Radiol Anat. 1994;16:31-6.
14. Ely MR, Cheuvront SN, Montain SJ. Neither cloud cover nor low solar loads are associated with fast marathon performance.
Med Sci Sport Exerc. 2007;39(11):2029-35.
15. Ely MR, Cheuvront SN, Roberts WO, Montain SJ. The impact of weather on marathon-running performance.
Med Sci Sport Exerc. 2007;39(3):487-93.
16. Florida-James G, Donaldson K, Stone V. Athens 2004: the pollution climate and athletic performance.
J Sport Sci. 2004;22:967-80.
17. Folinsbee LJ, Horvath SM, Raven PB, et al. Influence of exercise and heat stress on pulmonary function during ozone exposure.
J Appl Physiol. 1977;43:409-13.
18. Folinsbee LJ, Silverman F, Shephard RJ. Decrease of maximum work performance following O
3 exposure.
J Appl Physiol. 1977;42:531-6.
19. Gates MC. The influence of air pollution on thoroughbred race performance.
Equine Comp Exerc Physiol. 2007;4:79-88.
20. Gibbons SI, Adams WC. Combined effects of ozone exposure and ambient heat on exercising females.
J Appl Physiol. 1984;57:450-6.
21. Gliner JA, Horvath SM, Folinsbee LJ. Preexposure to low ozone concentration does not diminish the pulmonary function response on exposure to higher ozone concentrations.
Am Rev Respir Dis. 1983;127:51-5.
22. Hao J, Wang L. Improving urban air quality in China: Beijing case study.
J Air Waste Manag Assoc. 2005;55:1298-305.
23. Hazucha M, Silverman F, Parent C, Field S, Bates DV. Pulmonary function in man after short-term exposure to ozone.
Arch Environ Health. 1973;27:183-8.
24. Hu HC, Ben-Jebria A, Ultman JS. Longitudinal distribution of ozone in the lung: effects of respiratory flow.
J Appl Physiol. 1994;77:574-83.
25. Kabel JR, Ben-Jebria A, Ultman JS. Longitudinal distribution of ozone absorption in the lung: comparison of nasal and oral quiet breathing.
J Appl Physiol. 1994;77:2584-92.
26. Kesavanathan J, Swift DL. Human nasal passage particle deposition: the effect of particle size, flow rate, and anatomical factors.
Aerosol Sci Technol. 1998;28:457-63.
27. Kim CS, Hu SC. Regional deposition of inhaled particles in human lungs: comparison between men and women.
J Appl Physiol. 1998;84:1834-44.
28. Kinney PL, Thurston GD, Raizenne M. The effects of ambient ozone on lung function in children: a reanalysis of six summer camp studies.
Environ Health Perspect. 1996;104:170-4.
29. Linn WS, Venet TG, Shamoo DA, et al. Respiratory effects of sulfur dioxide in heavily exercising asthmatics.
Am Rev Respir Dis. 1983;127:278-83.
30. Liu Y, Matida EA, Gu J, Johnson MR. Numerical simulation of aerosol deposition in a 3-D human nasal cavity using RANS, RANS/EIM, and LES.
J Aerosol Sci. 2007;38:683-700.
31. Martin TR, Castle RG, Fredberg JJ, Wohl MB, Mead J. Airway size is related to sex but not lung size in normal adults.
J Appl Physiol. 1987;63:2042-7.
32. McDonnell WF, Smith MV. Description of acute ozone response as a function of exposure rate and total inhaled dose.
J Appl Physiol. 1994;76:2778-84.
33. Salthammer T. Heavy air pollution in Beijing and possible impact on Olympic athletes.
Clean Soil Air Water. 2008;36:731-3.
34. Shima M, Adachi M. Effect of outdoor and indoor nitrogen dioxide on respiratory symptoms in school children.
Int J Epidemiol. 2000;29:862-70.
35. Sun Y, Zhuang G, Wang Y, et al. The air-borne particulate pollution in Beijing-concentration, composition, distribution and sources.
Atmos Environ. 2004;38:5991-6004.
36. Thaller EI, Petronella SA, Hochman D, Howard S, Chhikara RS, Brooks EG. Moderate increases in ambient PM
2.5 and ozone are associated with lung function decreases in beach lifeguards.
J Occup Environ Med. 2008;50:202-11.
37. Wayne WS, Wehrle PF, Carroll RE. Oxidant air pollution and athletic performance.
J Am Med Assoc. 1967;199:151-4.
38. Yaglou CP, Minard D. Control of heat casualties at military training centers.
AMA Arch Ind Health. 1957;16:302-16.
