AIVC 11970 from exposure to infections, indoor air irritants, acute tained information on the dates and main causes of changes in weather, and antigens (e.g., pollen, house hospitalization. The data covered all the municipal mites, organic compounds).12•17 hospitals and Helsinki University Central Hospital, Ambient air pollution generally contributes to an in where more than 95% of patients with asthma who re creased frequency of asthma attacks, uHo and separate quired hospitalization were treated. This study includ correlations have been observed with increased concen ed only patients that were hospitalized primarily for trations of S02, nitrogen oxides (NO,), suspended partic acute asthma. Patients who were admitted to emergen ulates, 03, and other photochemical oxidants.10•21•25 cy wards and who needed more effective treatment The concentrations at which ambient air pollutants were also treated subsequently in bed wards, but they increase the frequency of asthma attacks at levels were studied separately. observed with epidemiological methods are not un Diagnosis of asthma was based on the World Health Asthma equivocally known. Nor is there adequate information Organization's Ninth International Classification of Dis and Low Level Air Pollution 1n Helsinki on the synergistic effects with various pollutants and eases (ICD-9). Only patients with bronchial asthma, di other factors,' e.g., cold, relative importance of the dif agnosis number 493 of the classification, were includ ferent causative factors, and differences in sensitivity of ed; therefore, those with chronic bronchitis were ex different age groups. The result of many studies have cluded. In Finland, diagnosis of asthma is always based been impaired by small sample sizes, confounding so on the consultation and statement of a specialist, which cioeconomic factors, insufficiently documented pollu are needed for the provision of medication. The state ANTII PONKA, M.D. tant levels, and biases in collecting information about ment must include a thorough history of the disease, Department of Environmental Health illnesses. findings of a clinical examination, and results of blood Helsinki City Health Department In Helsinki, where there is a population of 0.5 tests and pulmonary function tests. An increase in the Helsinki, Finland million, we investigated the incidence of asthma at peak expiratory flow of 15% or more by broncho tacks that required treatment in hospital wards. This in dilatating drugs, a decrease of at least 15% after exer cidence was correlated with ambient air SO,, NO, cise, or spontaneous changes in the peak expiratory NO,, CO, 03, TSP concentrations, temperature, rela flow of at least 20% within 6 h is needed as a criterion ABSTRACT. The effects tive humidity, and wind speed. of asthma diagnosis among adults and children who are of relatively low levels of air pollution and weather conditions on the number of patients who old enough to cooperate in peak expiratory flow meas had asthma attacks and who were admitted to a hospital were studied in Helsinki during urements. For each patient who visits a hospital ward a 3-y period. The number of admissions increased during cold Materials and methods w ather (n = 4 209), especially � among persons who were of working age but not among or an outpatient department, documents are com children. Even after standardization for temperature, all admissions, including emergency Air pollutants and meteorologic variables. Air pollu pleted according to a certain format. These are always ward admissions, were significantly correlated with ambient air concentrations of nitrogen tion measurements are conducted in Helsinki by dis checked, as is the diagnosis, by a senior specialist. dioxide (NO,), nitric oxide (NO), sulfur dioxide (S01), carbon monoxide (CO), ozone (03), trict municipal authorities. Sulfur dioxide is measured Statistical methods. The number of patients admitted and total suspended particulates (TSP). Regression analysis revealed that NO and 03 were by coulometric instruments at four automatic monitor each day was calculated and correlated with the cor most strongly associated with asthma problems. Effects of air pollutants and cold were ing stations, NO.x by chemiluminescence at two sta maximal responding mean daily concentrations of S02, NO,, if they occurred on the same day, except for 03, which had a more pronounced tions, CO by nondispersive infrared spectrometry at NO, CO, TSP, o., temperature, wind speed, humidity, effect after a 1-d lag. The associations between pollutants, low temperature, and admis two stations, and o. by ultraviolet absorption at one sions were most significant among adults of working and minimum hourly temperature during the day. _ age, followed by the elderly. Among children, only O, and NO station. Total suspended particulates are collected by Cold affects the incidence of asthma; therefore, par were significantly correlated with admissions. Levels of pollutants were fairly low, the high-volume samplers at six stations. At one weather tial correlations were also calculated after standardiza long-term mE:an being 19.2 µg/m3 for SO,, 38.6 µg/m3 for NO,, 22.0 µg/m'or O,, and station, measurements of temperature, wind speed, 1.3 mg/m' for CO. In contrast, the mean concentration of TSP tion for minimum temperature. The minimum tempera was high (76.3 µg/m'), and the mean and relative humidity are recorded hourly. temperature was low ( + 4.7 °C). These results sug ture was standardized because it was more strongly gest that concentrations of pollutants lower than those given as guidelines in many coun Helsinki City includes a relatively small area, i.e., 185 correlated with the frequency of asthma attacks than tries may increase the incidence 2 of asthma attacks. km • The main sources of air pollutants are energy pro was daily mean temperature. Correlations were also duction by coal-fired and oil-fired power plants, road calculated for 1- and 2-d lags. The log-transformed val traffic, and, to a small extent, industrialization. In 1987, ues of the variables were also used in the analysis. the total emission of N02 was 17 600 tons, of which Correlations were calculated separately for persons 32% was derived from traffic, 67% from energy pro who were 0-14, 15-64, and 65+ y of age. The relative EXPERIMENTAL AND EPIDEMIOLOGICAL STUDIES ing exercise.3·5 Less ambiguous results have been re duction, and 1% from industry. Ninety-seven percent importance of the various factors26•27 was estimated by have revealed that ambient air pollution and cold ported in studies of NO, exposure. Orehek et al.6 re of CO is emitted from cars. The monitoring stations that stepwise regression. weather increase the frequency of acute asthmatic epi ported potentiation of the carbachol bronchoconstric measure NO.x and CO are located on the main streets, The population in Helsinki numbered 488 604, sodes. The results of experimental studies have been tor response after exposure to 188 µg/m3 (0.1 ppm) where the bulk of NOx is produced by traffic. Emissions 491 148, and 491 777 in 1987, 1988, and 1989, respec more consistent than have those of epidemiological NO, in 13 of 20 asthmatics. Bronchoconstriction and of S02 total 22 500 tons, of which 93% is from energy tively. studies. airway hyperreactivity to cold air have been observed production, 2% from traffic, and 5% from industries. Asthmatics are adversely affected by many air pollu 3 after exposure to 560 µg/m (0.3 ppm) N02.2•5 Exposure The stacks at the power plants are 100- to 150-m high, tants and by cold weather. Bronchoconstriction occurs to ozone at concentrations as low as 234 µg/m3 (0.12 Results S02 whereas the exhaust gases from cars spread at street in certain asthmatic subjects upon exposure to ppm) increased airway reactivity in healthy subjects.2•7 level. At street level, 60-80% of NOx is derived from Number of admissions. During the 3-y period, 4 209 levels as low as 286 µg!m• (0.1 ppm) for 10 min during The frequency and intensity of asthmatic episodes exercise.1 In most studies, it has been shown that traffic. Therefore, NO.x and CO are indicators of pollu hospitalizations for asthma occurred, i.e., average of are related to season, e.g., autumn and frequency in tion resulting from traffic, and 502 is an indicator of 3.84 admissions/d. More than half (2 414) were admit changes in pulmonary function occur when concentra creases during the winter in the northern hemisphere. 3 pollution generated by energy production. ted by the emergency ward. In 1987, the number of ad tions of S02 are approximately 1 430 µg/m (0.5 ppm).2 The reasons for a seasonal relationship are not clear.a-11 Incidence of admissions for asthma. Data concern missions was 1 534; in 1988, 1 484; and in 1989, 1 221. Cold weather, either alone or in combination with S02 Perhaps the reasons are multifactorial and vary with ing hospital admissions for asthma attacks were obtain There were 1 359 cases who were in the 0-14 y age or NO,, causes and potentiates bronchospasm and air climate, vegetation, ambient air pollution, and other ed from the register that documented all periods of il group, 1 685 cases were 15-64 y of age, and 1 165 way hyperreactivity among asthmatics, especially dur- factors. Seasonal occurrence of asthma may also result lness that required hospitalization. The register con- were at least 65 y of age.
262 Archivesof Environmental Health September/October 1991 [Vol. 46 (No. 5)] 263
. I
__ l_ was only slightly less if there was a 2-d lag (Table 5). The 3.-Correlation Coefficients between Admissions for was also observed after these Table 1.-Mean Concentrations of Ambient Air Pollutants and Mean Values of Table effect of cold weather and Mean Daily Pollutant Concentrations and Metereological Variables in Helsinki, 1987-1989 Asthma lags. Temperatures Relative humidity had no effect on numbers of cases. �\ admissions However, wind speed was correlated with Variable Mean Range SD Admissions by � for asthma on the same day (p = .036) and after a 1-d All admissions emergency wards lag. 502 (µg/m2) 19.2 0.2-94.6 12.6 I of all other pollutants, except 03, N02 (µg/m3) 38.6 4.0-169.6 16.3 Concentrations through Friday than on Satur CO (mg/m2) 1.3 0-7.0 0.8 I so, .0926 .1319 were higher on Monday O, (µg/m3) 22.0 0-89.9 13.1 (.0022) (<.0001) day and Sunday. Therefore, correlations were calculat TSP (µg/m3) 76.3 6.0-414.0 51.6 ed separately for weekdays and for w�e��nds. Dur�ng Mean temperature ( 0 () 4.7 -37.0-+26.4 9.3 NO .2128 .1213 weekends the correlations were not s1gnif1cant; during Minimum daily temperature ( 0 () 2.4 -39-+24.0 9.3 (<.0001) (<.0001) of cases were associated Relative humidity (%) 82.9 37.4-100.0 12.0 weekdays: however, numbers .1213 of Wind speed (m/s) 4.7 0.6-10.9 1.8 N02 .1774 with NO, NO,,, and 03. The mean daily number (<.0001) (<.0001) cases during weekends was 2.44 and was 4.40 for Mon . co .1578 . 0650 day through Friday (<.0001) (.0335) Correlations with age. The relationship between ad Table 2.-Pearson Correlation Coefficients between Daily Mean Concentrations of missions for asthma and pollutants and cold weather .0074 .0739 Pollutants and Temperatures O, > 65 +-y age group (.8119) (.0170) was as follows: 15-64-y age group > 0- 14-y age group. . .0919 .1075 Minimum TSP Among children, only 03 and NO correlated with ad (.0136) (.0039) temperature S02 N02 NO o, TSP missions by emergency wards. Ozone, NO, and CO Mean temperature -.0541 -.0915 were also correlated with all admissions at the p < .OS Mean .9867 -.5507 -.1407 -.3043 .1559 -.0256 (.0841) (.0034) level. Among those who were 15-64 y of age, the situa temperature (<.0001) (<.0001) (<.0001) (<.0001) (<.0001) (.5089) tion was reversed: all pollutants, except 03, and low Minimum tem�rature -.0592 -.1006 Minimum 1 -.5712 -.1701 -.3162 .1504 -.0363 (.0585) (.0013) temperature contributed to an increase in number of temperature (<.0001) (<.0001) (<.0001) (<.0001) (.3490) cases. The effect of gaseous pollutants was greater than S02 1 .4516 .4773 -.1778 .1919 was the effect of TSP. Among the elderly, S02, NO, Note: p values appear in parentheses. (<.0001) (<.0001) (<.0001) (<.0001) N02, and cold weather correlated significantly with ad- NO, 1 .6664 -.2582 .1962 (<.0001) (<.0001) (<.00Q1) NO 1 -.5479 .1097 (<.0001) (.0034) o, 1 .1836 N 70 (<.0001) so, 90 Note: p values appear in parentheses. NO, -- 7i 60 '1 TSP ····•l·I f \ 80 I I .� I I ;. I I : : Pollutants and meteorological variables. Concentra Relation of air pollutants and weather to admis I \ I : : 50 I I /I 70 tions of most pollutants in the ambient air were rela sions. The frequency of all admissions for asthma was I \I I I \f I ·. highly significantly correlated with daily concentrations I tively low. The mean concentrations (presented as the I '·] I ·� 60 mean of the means at various stations) of 502, N02, of S02, NO, N0 , and CO and with the minimum hour I : I 2 , 3 , I CO, and 03 during the 3-y period were 19.2 µg/m , ly temperature. Frequency of admissions was also sig 40 , I , I 3 ,.. 38.6 µg/m3, 1.3 mg/m3, and 22.0 µg/m , respectively. nificantly correlated with TSP, but it was not correlated 50 The mean concentration of TSP was 76.3 1-1g/m3, which with concentration of 03 or the mean daily tempera was a high value that resulted from the (a) meteorolog ture. The number of asthma cases seen in the emergen JO 4 0 ical conditions; (b) the use of studded tires during cy wards correlated at least on the p - .OS level with all winter, whkh caused erosion of street surfaces; and (c) pollution variables and with cold weather (Table 3). JO the use of sand on streets to treat icy surfaces. Mean The weekly mean concentrations of 502, N01, and 20 values and ranges for the various variables are present TSP, and the number of asthma cases seen each week ed in Table 1. When we compared the different sta for 3 y are presented in Figures 1-3. 20 tions, the mean long-term concentrations of S02 had a A better estimate of the significance of the air pollu \ ../ .. 1.4 to 2.3-fold variation; those of TSP, a 2.6 to 2.8-fold tants was desired; therefore, partial correlations of I 0 0 variation; and those of N01, a 1.3 to 1.6-fold variation. these pollutants and admissions for asthma were also I Daily values of the different variables and the hourly calculated after standardization for temperature. Even minimum temperature were correlated (Table 2). High after standardization, the number of cases seeri at the ly significant negative correlations were obtained when emergency wards was positively correlated with all pol 01/11/87 01/01/88 02/0J/87 02/05/87 02/07/87 01/09/87 S01, N02, and NO were compared with temperature, lutants, except CO, as was the number of all cases with Jl/12/86 whereas 03 correlated positively with temperature. all pollutants, except 03 (Table 4). Year 1987 There was a highly significant correlation between S02 Except for 03, the significance of the correlations and NO, N02, and TSP, but S01 was correlated in disappeared after a 1- or 2-d lag . The effect of 03 on · of admissions that resulted from asthma (thick line) in 1987. Fig. 1. Weekly mean concentrations of 5021 N02, and TSP, and numbers versely with 03. numbers of cases was greatest if there was a 1-d lag and 265 September/October 1991 [Vol.46 (No. 5)] 264 Archivesof Environmental Health -·
100 g;m3 were also compared during the highest quartile of air SO, µ Table 4.-Partial Correlations between Admissions for Asthma and Mean Daily Pollutant Concentrations After pollutants and during the three lowest quartiles (Table NO, .. .:.. 90 Standardization for Minimum Temperature 7). During the period of high S02 pollution, the mean 60 number of all admissions was 7% greater than during TSP /\ the period of lower pollution. During periods of high 80 Admissions by All admissions emergency wards NO, N02, CO, and TSP pollution, the mean number of 50 all admissions was 28%, 29%, 15%, and 18%, respec 70 tively, greater than during the period of lower pollu so, .0770 .1050 (.0172) (.0011) tion. 60 NO .2054 .1664 40 (<.0001) (<.0001) Discussion N02 .1830 .1 137 50 This study suggests that ambient air pollutants in (<.0001) ( . 0004) crease asthma attacks at lower concentrations than 30 co .1426 .0391 4 0 (<.0001) (.2273) previously believed. However, the results of earlier epi O, .0289 .1 083 demiological studies are inconsistent. The cold climate (.3725) (.0008) also increased the effects of pollutants on the airways. 20 30 TSP .0875 .0995 No clear-cut limits can be provided on the basis of (.0301) (.0137) health effects, but these increase gradually as concen 2 0 trations of pollutants increase. Perhaps the comprehen Note: 10 p values appear in parentheses. sive collection of data for a large population over a long 10 period, as was done in the present study, can reveal the health effects more precisely than did some earlier missions to emergency wards, and CO correlated sig studies that were completed with more limited material. 0 nificantly with all asthma cases. In contrast, 03 had no Several studies have reported that frequency of asth 31/12/87 01/03/88 01/05/88 01/07/88 31/08/88 ll/10/88 31/12/88 effect on asthma among the elderly. ma attacks increases at low levels of pollution. Cohen Year 1988 Separation of the effects of weather and pollution et al.,21 in a follow-up of 20 asthmatics for 7 mo, ob variables. Temperature was significantly associated served significant correlations between attack rates and
Fig. 2. Weekly mean concentrations of SO,, NO,, and TSP, and numbers of admissions that resulted from asthma (thick line) in 1988. with admissions; therefore, we examined the effects of temperature and between attack rates and pollution pollutants by using multiple-regression analysis after levels (after temperature was standardized). The most standardization of minimum temperature. In the step important single factor was cold weather. When (a) the wise analysis, NO was the most strongly associated long-term S02 concentration exceeded 200 µg/m3 3 N 70 with all admissions (p < .0001), followed by 03 (p < (0.07 ppm), (b) TSP exceeded 150 µg/m , or (c) 100 m3 SO, µg; .0001). The association with CO was almost significant temperature was less than 32 °F, the increase in asthma :: (p = .038). The effects of N02 and S02 were at least attacks was significant. If the wind speed exceeded 4 NO, 90 partially masked by NO. If the model included only mph or the relative humidity was less than 80%, the in 6 0 . TSP N02, S02, 03, and CO, the order of significance was as crease was almost significant . 80 follows: N02 (p < .0001), CO (p < .0001), 03 (p = Goren and Hellmann25 investigated the prevalence of .006), and S02 (not significant). The explanatory power asthma and respiratory symptoms among schoolchil .. 2 50 f , of the model was low (r = 0.093), but it increased dren who lived in a more polluted area and a less : : 70 ="°\ slightly when log-transformed values of the variables polluted area in Israel. The relative risk of asthma was 2 were used (r = .139). 2.66-fold in the more polluted area. The mean monthly so 40 Comparison of admissions during periods of high S02 and NO,, concentrations in the more polluted area and low pollution. The number of admissions per day did not exceed 11-45 µg/m3 and 8-33 µg/m3, respec-
50
30 40 Table 5.-Correlations between Admissions for Asthma and Mean Daily Concentrations of Ozone and Temperatures Simultaneously and wi!h 1- and 2-d Lags
20 30 Same day 1-d lag 2-d lag
20 Admissions by Admissions by Admissions by All admissions emergency wards All admissions emergency wards All admissions emergency wards I 0
1 0 Ozone .0074 .0739 .1 499 .1581 .1407 .1509 (NS) (.0170) (<.0001) (<.0001) (<.0001) (<.0001) Mean daily -.00541 -.0915 -.0503 -.0821 -.0400 -.0778 temperature (0.0841) (.0034) (NS) (.0087) (NS) (.0130) 31/12/88 02/0l/89 02/05/89 02/07/89 01/09/89 01/11/89 01/01/90 Minimum temperature -.0592 -.1 006 -.0586 -.091 1 -.0407 -.0822 Year 1989 (.0585) (.0013) (.0612) (.0036) (NS) (.0087)
Note: p values that exceeded .10 are provided in parentheses. Fig. 3. Weekly mean concentrations of S02, N02, and TSP, and numbers of admissions that resulted from asthma (thick line) in 1989.
266 Archives of Environmental Health September/October 1991 [Vol. 46 (No. 5)] 267 " .
·-
nual concentrations in the polluted area were 120-570 2. Air Quality Guidelines for Europe. WHO Regional Publications, Table 6.-Correlations between Admissions for Asthma and Pollutants and Temperatures, by Age Group European series no. 23. World Health Organization, µg/m3 for NO, 80-280 µg/m3 for N02, and 2.0-9.0 Copenhagen; 1987. mg/m3 for CO. In contrast, S02 levels of 120-140 µg/m3 3. Bethel RA, Sheppard D, Epstein J, Tam E, Nadel JA, Boushey HA. 0-14 y 15-64 y > 64 y and 0 levels of 70-130 µg/m3 did not have any effect. 3 Interaction of sulfur dioxide and dry cold air in causing bron Admissions by Admissions by Admissions by Unlike the above-mentioned studies, some other choconstriction in asthmatic subjects. Appl Physiol 1984; 57: All admissions emergency wards All admissions emergency wards All admissions emergency wards studies have not reported associations between the in 419-23. 4. Linn WS, Shamoo DA, Vinet TG, et al. Combined effect of sulfur cidence of asthma attacks and air pollutants in similar dioxide and cold in exercising asthmatics. Arch Environ Health so, -.01391 .0332 .1039 .1199 .0796 .1169 contentrations. 2�32 This unambiguous finding cannot be 1984; 39:339-46. (NS) (NS) (.0006) (<.0001) (.0085) (<.0001) explained by the cold weather, although cold is associat 5. Bauer MA, Utell MJ, Morrow PE, Speers DM, Gibb FR. Inhalation NO .0737 .0772 .0170 .1543 .1557 .1419 ed with an increased incidence of asthma attacks. 21 •22•31•33 of 0.30 ppm nitrogen dioxide potentiates exercise-induced bron (.0155) (.0112) (<.0001) (<.0001) (<.0001) (<.0001) chospasm in asthmatics. Am Rev Respir Dis 1986; 134:1203-08. Even though a similar correlation was observed in the N02 .0166 .0061 .1648 .1189 .1501 .1392 6. Orehek J, Massari JP, Gayrard P, Grimaud C, Charpin J. Effect of present study between the frequency of hospital admis (NS) (NS) (<.0001) (<.0001) (<.0001) (<.0001) short-term, low-level nitrogen dioxide exposure on bronchial sions for asthma attacks and air pollution, it must be sensitivity of asthmatic patients. J Clin Invest 1976; 57:301-07. co .0747 .0107 .1257 .0906 .0913 .0303 Folinsbee LJ, McDonnell WF, Horstman DH. Pulmonary function (.0146) (NS) (<.0001) (.0030) (.0028) (.3223) remembered that the incidence of asthma as a disease 7. was not considered in this study; rather, frequency of and symptom responses after 6.6-hour exposure to 0.12 ppm O, .0658 .1051 -.0480 -.0091 .0061 .0203 acute attacks among persons with bronchial asthma ozone with moderate exercise. J Air Pollut Control Assoc 1988; (.0336) (.0007) (NS) (NS) (NS) (NS) 38:28-35. TSP .0489 .0592 .0685 .0775 .0530 .0686 was noted. 8. Booth S, DeGroot J, Markusa R, Horton RJM. Detection of (NS) (NS) (.0662) (.0376) (NS) (.0656) The harmful effects of S02, TSP, and N02 on asthma epidemics in seven cities. Arch Environ Health 1965; Mean daily .0102 .0011 -.0604 -.1082 -.0506 -.0913 asthmatic persons are generally known. Conversely, 10:152-55. 9. Goldstein JF, Currie B. Seasonal patterns of asthma: a clue to temperature (NS) (NS) (.0538) (.0005) (NS) (.0035) NO and CO are not toxic to airways. The correlations etiology. Environ Res 1984; 33:201-15. Minimum temperature observed in our study suggest that these compounds .0107 -.0040 -.0721 -.1165 -.0467 -.0948 10. Bates DV, Baker-Anderson M, Sizto R. Asthma attack periodicity: (NS) (NS) (.0213) (.0002) (NS) (.0024) are indicators of air pollution, especially pollution that a study of hospital emergency visits in Vancouver. Environ Res results from traffic. 1990; 51 :51-70. Note: p values that exceeded .10 are provided in parentheses. The frequency of admissions by emergency wards 11. Weiss KB. Seasonal trends in US asthma hospitalizations and mortality. JAMA 1990; 263:2323-28. was more highly correlated with S02, 0 , TSP, and cold 3 12. Clarke CW. Relationship of bacterial and viral infections to ex weather than was frequency of all admissions, but all acerbations of asthma. Thorax 1979; 34:344-47. admissions correlated slightly better with pollutants Table 7.-Mean Daily Number of Admissions During the Low and High Pollution Periods 13. Murray AB, Ferguson AC, Morrison B. The seasonal variation of that primarily indicated traffic pollution on the' street allergic respiratory symptoms induced by house dust mites. Ann level, i.e., NO, N02, and CO. Persons with severe Allergy 1980; 45:347-50. Admissions by Mean concentration symptoms were first treated at the emergency wards 14. Packe GE, Ayres JG. Asthma outbreak during a thunderstorm. All admissions emergency wards of pollutants Lancet 1985; i:199-204. and were later transferred to ordinary bed wards; 15. Lebowitz MD, Colins L, Holberg CJ. Time series analyses of therefore, the difference may have resulted from more respiratory responses to indoor and outdoor environmental so, Low 3.78 2.08 11.0 µg/m3 severe reactions in the respiratory tract caused by S02, phenomena. Environ Res 1987; 43:332-41. 23.1 High 4.05 2.57 16. Platt-Mills TAE, Hayden ML, Chapman MD, Wilkins SR. Seasonal p TSP, 0 , and cold weather among certain groups of NS .0002 3 variation in dust mite and grass-pollen allergens in dust from the sensitive asthmatic persons. However, the better corre NO Low 3.60 2.01 49.0 µg/m3 houses of patients with asthma. J Allergy Clin lmmunol 1987; High 4.59 2.72 116.3 lation of all admissions with NO, N02, and CO may 79:781-91. p <.0001 <.0001 have reflected a more comprehensive exposure of the 17. Anto JM, Sunyer J, Rodriquez-Raisin R, Suarez-Cervera M, Vaz N02 Low 3.63 2.06 28.1 µg/m3 population to traffic pollution when compared with quez L, and the Toxicoepidemiological Committee. Community High 4.59 2.58 45.8 that caused by energy production. outbreaks of asthma associated with inhalation of soybean dust. N Engl J Med 1989; 320:1097-1102. p <.0001 <.0001 Despite that fact that the association of incidence of 18. Shrenk HH, Heimann H, Clayton GD, Gafafer WM, Wexler H. co Low 3.73 2.18 0.8 mg/m> asthma attacks with relatively low levels of pollutants Air pollution in Donora, epidemiology of the unusual smog 1.7 High 4.27 2.23 and cold weather was obvious in our study, these fac episode of October 1948: preliminary report. Public Health p .0053 NS tors account for an explanatory power of approximate Bulletin No. 306, Washington, DC, 1949. o, Low 3.58 2.15 12.3 µg/m3 19. Logan WPD. Mortality in the London fog incident. Lancet 1953; ly 14 % in regression analysis. Other factors were proba High 3.84 2.28 29.4 i:336-38. p NS NS bly more important. 20. Ministry of Health. Morbidity and mortality during the London The results of the present study, and those of some TSP Low 3.64 2.06 42.3 µg/m3 fog. Rep Pub Health Med Subj, No. 95. London: Her Majesty's High 4.28 2.57 93.1 other studies, show that ambient air pollutants, at con Station office, 1954. AA, p .0030 .0018 centrations lower than the guidelines provided in many 21. Cohen Bromberg S, Buechley RW, Heiderscheit MA, Shy CM. Asthma and air pollution from a coal-fueled power plant. countries and by the World Health Organization/ may Am J Public Health 1972; 62:1181-88. cause an increase in the incidence of asthma. 22. Whittemore AS, Korn EL. Asthma and air pollution in the Los tively, but the maximum half-hour concentrations were S02 but not with N02 or 0 levels. This phenomenon Angeles area. Am J Public Health 1980; 70:687-96. 3 * * * • • * * * * • rather high at 133-836 µg/m3 and 38-528 µg/m3, was observed in the 15-60-y age group during the sum 23. Ussetti P, Roca J, Agusti AGN, Montserrat JM, Rodriquez-Raisin Submitted for publication August 27, 1990; revised; accepted for respectively. mer and in the over 61-y age group during the winter. R, Agusti-Vidal A. Another asthma outbreak in Barcelona: role of publication January 21, 1991. oxides of nitrogen. Lancet 1984; i:156. In Los Angeles, 'Whittemore and Korn22 found a This association was observed during the same day or if Requests for reprints should be sent to: Antti Ponka, M.D., Chief, 24. Imai M, Yoshida K, Kitabatake M. Mortality from asthma and higher attack rate among adult asthmatics if the con there was a 1-d lag. Exact concentrations of pollutants Department of Environmental Health, Helsinki City Health Depart chronic bronchitis associated with changes in sulfur oxides air centrations of oxidants and particulates were high or if were not provided, but the highest hourly concentra ment, Viipurinkatu 2,. 00510 Helsinki, Finland. pollution. Arch Environ Health 1986; 41:29-35. 25. Goren AJ, Hellmann S. Prevalence of respiratory symptoms and the temperature low. The medians of 8-mo mean con tions of S02, N02, and 0 reached 286-572 µg/m3 * * * * * * * * * * 3 diseases in schoolchildren living in a polluted and in a low centrations of particulates and oxidants were 51-121 (0.1-0.2 ppm), 564-1 128 µg/m3 (0.3-0.6 ppm), and polluted area in Israel. Environ Res 1988; 45:28-37. References µg/m3 and 0.03-0.15 ppm, respectively. 400-800 µg/m3 (0.2-0.4 ppm), respectively. 26. Draper NR, Smith H. Applied regression analysis, 2nd ed. New Bates et al. 10 observed that the number of atten A German questionnaire survey28 revealed that among 1. Sheppard D, Saisho A, Nadel JA, Boushey HA. Exercise increases York: John Wiley & Sons, 1981. dances that resulted from asthma at the emergency children, asthma was more common in areas where sulfur dioxide-induced bronchoconstriction in asthmatic sub 27. SAS user's guide: statistics, 5th ed. Cary, NC: SAS Institute Inc., departments of Vancouver hospitals correlated with there were high NO, N02, and CO levels. The mean an- jects. Am Rev Respir Dis 1981; 123:486-91. 1985.
268 Archives of Environmental Health September/October 1991 [Vol. 46(No. 5)] 269 '·.
28. Wichmann H-E, Hubner HR, Malin E, et al. The relevance of 40:332-45. health risks by ambient air pollution, demonstrated by a cross 31. Vedal S, Schenker MB, Munoz A, Samet JM, Batterman S, Speizer section study on croup syndrome in Baden-Wurttenberg. Off TE. Daily air pollution effects on children's respiratory symptoms Gesundh 1989; 51:414-20 (in German). and peak expiratory flow. Am J Public Health 1987; 77:694-98. 29. Perry CB, Chai H, Dickey OW, et al. Effects of particulate air 32. Berciano FA, Dominguez J, Alvarez FV. Influence of air pollution pollution on asthmatics. Am J Public Health 1983; 73:50-56. on extrinsic childhood asthma. Ann Allergy 1989; 62:135-41. 30. Goldstein JF, Weinstein AL. Air pollution and asthma: effect of 33. Fleischer SLM, Asnani GC. The influence of weather on asthma in exposure to short-term sulfur dioxide peaks. Envirion Res 1986; Nairobi. Int J Biometeor 1978; 22:263-70.
Ambient Air Pollution and Cancer in
California Seventh-day Adventists
PAULK. MILLS, Ph.D., M.P.H. DAVID ABBEY, Ph.D. archives of W. LAW RENCE BEESON, M.S.P.H . • FLOYD PETERSEN, M.P. H. Department of Public Health and Environmental Health Preventive Medicine School of Medicine Loma Linda University
• • • • • • • • • Loma Linda, California • • • • • • • • • • • • • For 47 years, this noted ORDER FORM • • journal has provided ob- DYES! I would like to order a one-year subscription to Archives of Environ jective documentation of mental Health, published bimonthly. I understand payment canbe made to • the effects of environ- ABSTRACT. Cancer incidence and mortality in a cohort of 6 000 Seventh-day Adventist Heldref Publications or charged to my VISA/MasterCard (circle one). • nonsmokers who were residents of California were monitored for a 6-y period, and rela D mental agents on human $85.00 annual rate • tionships with long-term ambient concentrations of total suspended particulates (TSPs) ACCOUNH health. In one single ______EXPIR AT ION DA TE ______and ozone (03) were studied. Ambient concentrations were expressed as mean concentra
• source Archives of Envi- tions and exceedance frequencies, which are the number of hours during which concen SIGNATURE______• ronmental Health: An trations exceeded specified cutoffs (e.g., federal and California air quality standards). Risk of malignant neoplasms in females increased concurrently with exceedance frequencies • International Journal for all TSP cutoffs, except the lowest, and these increased risks were highly statistically sig NAME/INSTIT UTION·------brings together the latest nificant. An increased risk of respiratory cancers was associated with only one cutoff of O,, • and this result was of borderline significance. These results are presented in the context of ADDRESS ______• research from such vary- _ _ setting standards for these two air pollutants. • ing fields as epidemiol- CITYiSTATEiZl �'"------ogy, toxicology, biosta-
COUNTR Y______• • tistics, and biochemistry . ADD FOR POSTAGE OUTSIDE THE U.S. ALLOW WEEKS FOR DE LIVERY OF FIRST ISSUE. $12.00 6 • In a field where today's ADVERSE HEALTH EFFECTS associated with ambient cross-sectional or correlational design.5•6 In this ap SEND ORDER FORM AND PAYMENT TO: certainty often becomes air pollution have been scrutinized by environmental proach, indices of air pollution in certain geographic • tomorrow's myth, this HELDREF PUBLICATIONS, ARCHIVES OF ENVIRONMENTAL HEALTH • scientists for most of the twentieth century.1•2 There is locales were correlated with the age-adjusted cancer • journalpublishes new 1319 EIGHTEENTH STREET, NW, WASHINGTON, DC 20036-1802 evidence that air pollution contributes to morbidity mortality rates in the same areas, with or without ad research with scientific from airway obstructive disease and other forms of res justment for consumption of cigarettes. Most of these PHONE (202) 296·6267 FAX (202) 296-5149 • piratory disease, cancer, and cardiovascular disease.3•4 studies have not demonstrated that an increased can • integrity andrigorous SUBSCRIPTION ORDERS 1 (800) 365-9753 It is difficult to evaluate this evidence because the ef cer risk occurs with an increase in air pollution levels . • methodology. fects of tobacco smoke are difficult to separate from the In fact, the U.S. Environmental Protection Agency cri effects of air pollution, especially in urban areas where teria document for particulate matter and sulfur oxides the numbers of smokers and concentrations of air pol concluded: " ...nor does there presently exist credible lutants are higher than in more rural areas. Cigarette epidemiological evidence linking increased cancer smoking has been implicated in the etiology of cardio rates to elevations in particulate matter as a class, i.e., vascular disease, respiratory diseases, lung cancer, and undifferentiated as to chemical content."7 In many in several other forms of cancer in humans (e.g., cancers stances, however, an inability to detect relationships of the bladder and pancreas). can be linked with limitations in study design, poor Most studies of air pollution and cancer in humans statistical power, or other methodological weaknesses. have been epidemiologic investigations that used Seventh-day Adventists provide a unique opportunity
September/October 1991 [Vol. 46(No. 5)] 271 270 Archives of Environmental Health
--...... -- - *""-�