- Email: [email protected]
Influence of Atmospheric Pressure, Outdoor Temperature, and Weather Phases on the Onset of Spontaneous Pneumothorax
Influence of Atmospheric Pressure, Outdoor Temperature, and Weather Phases on the Onset of Spontaneous Pneumothorax BRANKA BULAJICH, MD, MR SCI, DRAGAN SUBOTICH, MD, PHD, FCCP, DRAGAN MANDARICH, MD, PHD, RADMILA VOJNOVICH KLJAJICH, BSC, AND MILAN GAJICH, BSC
PURPOSE: To analyze the influence of meteorological factors such as atmospheric pressure (AP), outdoor temperature (T) changes, and weather phases (WP) on the occurrence of spontaneous pneumothorax (SP). METHODS: Retrospective study, including 659 patients with primary SP and SP associated with chronic obstructive pulmonary disease (COPD), conservatively treated in a 5-year period. In the analyzed period, 548 days with SP were compared both with 3 days preceding the onset of SP and with days without pneumothorax. The comparison was made depending on weather phases and on different aspects of AP and T. RESULTS: Seasons of the year did not significantly influence the occurrence of SP. No significant difference was found between SP and non-SP days depending on different aspects of AP and T changes. Most patients were admitted in ‘‘clusters’’ with not more than 2, 3, or 4 days between two successive admissions. The occurrence of SP was significantly correlated with weather phases 2ts (anticyclonic situation with warm and dry weather) and 5 hv (passing of the cold front). Biological sense and possible explanations of this correlation are discussed. CONCLUSIONS: Among all analyzed meteorological factors, significant correlation was found only between weather phases 2ts and 5hv and the occurrence of SP. Ann Epidemiol 2005;15:185–190. Ó 2004 Elsevier Inc. All rights reserved. KEY WORDS:
Weather Phases, Pneumothorax, Atmospheric Pressure, Temperature.
INTRODUCTION The explanation of the influence of meteorological factors to the onset of spontaneous pneumothorax (SP) is the assumption that as long as free communication exists between subpleural air-containing cysts and atmospheric air, pneumothorax will not ensue, even in presence of major atmospheric pressure (AP) changes. Once this communication is disrupted, these cysts become isolated from the surrounding lung tissue, changing their morphology according to Boyle-Marriott’s law (P ! V Z const.), i.e., a fall in AP causes an increase in their volume (1). Although the morphology and ultrastructure of causative lesions in primary SP are well known, the reason for rupture of air-containing cysts is not absolutely clear (2). Broad
From the Intensive Care Unit (B.B.), and Clinic for Thoracic Surgery (D.S., D.M.), Institute for Lung Diseases, Clinical Center of Serbia, Belgrade, Serbia and Montenegro; Republic Hydrometeorological Service of Serbia, Belgrade, Serbia and Montenegro (R.V.K.); and Institute for Medical Statistics, Faculty of Medicine, Belgrade, Serbia and Montenegro (M.G.). Address correspondence to: Dragan Subotich, M.D., Ph.D., F.C.C.P., Clinic for Thoracic Surgery, Institute for Lung Diseases, Clinical Center of Serbia, Visegradska 26/20, Belgrade 11000, Serbia and Montenegro. Tel.: C381-11-361-5559; Fax: C381-11-646-988. E-mail: [email protected]; [email protected] Received December 15, 2003; accepted April 25, 2004. Ó 2004 Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010
consensus concerning the role of atmospheric factors in SP does not exist (3). The aim of this study, apart from the analysis of the influence of AP and outdoor temperature changes on the occurrence of SP, was to assess the relationship between SP and different weather phases in days with SP and during 3 preceding days. The term ‘‘weather phase’’ summarizes the influence of several meteorological factors, indicates the dynamics of their changes, thus directly correlates them with possible host response mechanisms (4, 5).
METHODS AND MATERIALS Patients The study included 659 patients with SP conservatively treated at the Institute for Lung Diseases in Belgrade between January 1990 and December 1994. Only patients with primary SP and SP associated with chronic obstructive pulmonary disease (COPD) were included. Of a total number of 659 admitted patients, 565 (85.7%) patients had primary SP and 94 (14.3%) patients had SP associated with COPD. There were 529 (80.3%) males and 130 (19.7%) females. Three hundred fifty (53.11%) patients belonged to the age group 21 to 40 years, whilst only 76 (11.54%) and 77 (11.68%) patients were younger than 20 years and older than 60 years, respectively. In 59.33%, SP was right sided. 1047-2797/05/$–see front matter doi:10.1016/j.annepidem.2004.04.006
186
Bulajich et al. ATMOSPHERIC FACTORS AND PNEUMOTHORAX
The criterion for the onset of SP was the day of the onset of symptoms reported by the patient. Only patients with the exact date of the onset of SP were included in the study. Also, only patients who contracted SP in the Belgrade region were included in the study, provided that they were there at least one day before the onset of symptoms. Meteorological Data The Republic Hydrometeorological Service in Belgrade provided atmospheric pressure (AP) and temperature (T) measurements. For each of the 1825 days in the analyzed 5year period, AP (mBar) measured at 7 hours, 14 hours, and 21 hours, as well as mean, maximal, and minimal AP values were registered. Temperature ( C) was measured at the same intervals, together with mean, maximal, and minimal T values. The aforementioned data were registered on the day of the onset of SP (‘‘day 0’’) and during 3 previous days (‘‘day ÿ 1,’’ ‘‘day ÿ 2,’’ and ‘‘day ÿ 3’’). Each day in the analyzed period corresponded to one of 20 weather phases (according to Hess-Berezowsky) that were registered and entered in the database. Names and main characteristics of weather phases analyzed in this study are as follows: 1hsdmild influence of anticyclone with cold and dry weather; 1tsdmild influence of anticyclone with warm and dry weather; 2hsdanticyclonic situation with cold and dry weather; 2tsdanticyclonic situation with warm and dry weather; 2tvdanticyclonic situation with warm and wet weather; 3atsdcenter of the anticyclone, warm and dry weather; 3ftsdcenter of the cyclone, warm and dry weather; 4hvdfront side of the cyclone with cold and wet weather; 4tsdwarm, dry weather, preceding the weather change; 4tvdfront side of the cyclone with warm and wet weather; 5 hs, 5 hv, 5 ts, 5tvdpassing of fronts (5hs: cold front; 5hv: occlusion, cold front type; 5ts: warm front; 5tv: occlusion, warm front type); 6 zhsdweather stabilization with cold and dry weather; 6zhvdweather stabilization with cold and wet weather; 6ztsddry weather between two fronts; 6ztvdwarm and wet weather preceding the cold front; 6hsdweather situation with temperature inversion; and 6hvdfog during the whole day. Data Analysis In the analyzed period there were 548 days with SP. These days were compared both with 3 days preceding the onset of SP and with all days without pneumothorax in the analyzed period. The comparison was made depending on different meteorological factorsdmean, maximal, and minimal AP and T values; maximal daily AP- and T-amplitudes; significant AP variations (AP amplitudes of at least 10 mBar in a 24-hour interval). For this type of the analysis, expected frequency of significant AP changes was calculated and compared with registered frequency. Days with
AEP Vol. 15, No. 3 March 2005: 185–190
pneumothorax were also compared with 3 preceding days depending on weather phases. Comparisons depending on different meteorological factors were made both with type of SP independently and depending on whether SP was primary or associated with COPD. Statistical analysis included distribution of frequencies, c2 test, t-test for paired samples, t-test for unpaired samples, and Wilcoxon-Mann-Whitney U-test.
RESULTS Seasons and the Occurrence of SP The frequency of SP occurrence was nearly equal in all seasons of the year. In spring, SP occurred in 171 (25.9%) and in summer in 163 (24.7%) patients; in most patients (174 [26.4%]) SP developed in autumn, whilst winter was the season with minimal (151 [23%]) number of patients with SP. The monthly distribution of admitted SP patients was different depending on the year. The most prominent variations were registered in February and Decemberdthe number of admitted patients in February 1992 was threefold the number for the same month in 1994; similar difference exists between December 1994 and December 1992. Less prominent variations were registered in April, June, and October. General Meteorological Features of SP-days and Non-SP Days On SP-days AP varied between 980.10 and 1023.50 mBar (mean value, 1002.19 G 7.21). Neither for AP nor for T were differences between SP-days and non-SP days statistically significant. Maximal daily AP amplitudes on SP-days were between 0.10 and 18.80 mBar (mean value, 2.89 G 2.32); on non-SP days, maximal amplitudes were between 0 and 16.10 mBar (mean value, 2.90 G 2.35). On SP-days, maximal daily T amplitudes were between ÿ 1.60 C and 19.80 C, mean value 9.46 G 3.98. Neither for AP nor for T were differences between SP-days and non-SP days statistically significant. Distribution of different AP intervals on SP-days and on non-SP days is presented in Table 1. Differences between SP-days and non SP-days depending on these intervals are not statistically significant. Similarly, differences between each group and the distribution in the entire analyzed period are not statistically significant. The analysis of the AP trend (stable, rising, or falling) on days with and without SP is also presented in Table 1. With AP change of at least 10 mBar accepted as significant, in 93.06% of days with SP there were no significant fluctuations in relation to the preceding day. Differences between SP-days and non-SP days are not statistically significant. With AP changes of at least 5 mBar as significant, although
AEP Vol. 15, No. 3 March 2005: 185–190
Bulajich et al. ATMOSPHERIC FACTORS AND PNEUMOTHORAX
TABLE 1. Atmospheric pressure intervals and trend on days with and without pneumothorax Days without SP
Days with SP
AP interval (mBar)
N
%
N
%
! 980 980–1000 1001–1010 1011–1020 1021–1030
102 522 470 154 29
7.98 40.87 36.80 12.05 2.30
47 214 221 61 5
8.57 39.05 40.32 11.13 0.93
1277
100
548
100
Total
Days without SP AP trend* Falling Stable Rising Total
Total N
149 8.16 736 40.33 691 37.86 215 11.78 34 1.87 1825
Days with SP
n
%
n
%
46 1206 25 1277
3.60 94.22 2.18 100
18 510 20 548
3.30 93.06 3.64 100
%
100
Total n
%
64 3.50 1716 94.02 45 2.48 1825 100
P O 0.05
187
significant AP-changes of patients with primary SP is identical to the exposure of SP-patients with COPD. The expected frequency of significant AP changes was calculated and compared with the observed frequency, both for the entire group and separately for primary SP and SP with COPD (Table 3). For the entire group, three or more than four exposures to such AP changes during 24 hours were significantly more frequent than expected. Outdoor Temperature Temperature differences between SP-days and non-SP days compared with the preceding day for both groups are presented in Table 4. Although in days preceding the occurrence of SP a certain temperature drop was registered (compared with non-SP days that were preceded by a day with a light temperature rise), no significant differences in mean temperature between SP-days vs. the preceding days were found.
*AP changes related to the preceding day (7-hour day – 1:7-hour day 0).
Weather Phases AP was stable in only 68.24% of SP-days (data not shown), differences between SP-days and non-SP days were not statistically significant.
Atmospheric Pressure Significant AP changes (DAP of at least 10 mBar in a 24hour interval) during the 3 days preceding SP are presented in Table 2. Of a total of 659 patients with SP, in 491 (74.65%) patients no significant AP changes were registered; the rest (?25%) of the patients were exposed to at least one, maximally to six, significant AP changes. Most of them (140 [16.23%]) were exposed to between one and three such episodes; 15 (2.30%) patients were exposed to four significant episodes, whilst five and six episodes were registered in 6 (0.91%) patients each. The exposure to TABLE 2. Significant AP changes (G 10 mBar) during three days preceding SP (24-hour interval)
No. of changes 0 1 2 3 4 5 6 Total
Primary SP
SP C COPD
n
%
n
%
n
%
420 35 46 38 14 6 6 565
74.30 6.20 8.20 6.70 2.50 1.10 1.10 100
72 3 8 10 1 0 0 94
76.60 3.20 8.50 10.60 1.10 0 0 100
492 38 54 48 15 6 6 659
74.60 5.80 8.20 7.30 2.30 0.90 0.90 100
P O 0.05
Total
In the analyzed period, SP occurred most frequently in presence of two weather phasesdphase 2ts and phase 5hv, each with 120 (21.9%) SP days (Fig. 1). The third most frequent was the weather phase 2hs with 11.9% SP days. Weather phase 6zhv was associated with 8.8% SP days. No differences in dominant weather phases distribution existed during the 3 days preceding SP. Concerning dominant combinations of weather phases, during the 2 days preceding SP, the most frequent pairs of weather phases were 2ts-2ts and 6zhv-5hv registered in 11 and 10 patients respectively. Neither of analyzed combinations was of statistical significance. Occurrence of Spontaneous Pneumothorax in Series (‘‘Clusters’’) In the analyzed 5-year period there were 548 (29.96%) days with SP. Most days (456 [83.21%]) had only one SP admission; on 75 (13.68%) and 16 (2.91%) days, two and three SP admissions were registered, respectively. Four SP admissions were registered on only one day during the entire period. TABLE 3. Comparison of the observed with the expected frequency of significant AP changes (24 hour interval) Number of exposures to significant AP changes (Os/Oc) 0 Primary SP 419/418 SP C COPD 72/70 All 491/488
1
2
35/51 3/9 38/60
46/45 8/7 54/52
Os/oc = observed frequency/expected frequency. *Statistically significant difference.
3 38/31 10/5 48/36*
4
O4
14/11 1/2 15/13
12/8 0/1 12/9*
188
Bulajich et al. ATMOSPHERIC FACTORS AND PNEUMOTHORAX
TABLE 4. Differences in the outdoor temperature between the ‘‘day 0’’ and the ‘‘day ÿ 1’’ on days with and without SP D T in relation to day ÿ 1 a
Mean D T1b D T2c
Days with SP
Days without SP
ÿ0.02 G 2.82 ÿ9.63 G 3.65 9.44 G 4.92
0.01 G 2.86 ÿ9.30 G 3.97 9.39 G 4.77
P O 0.05 a
Mean: difference in mean T between the day ÿ 1 and the day 0. D T1: difference between minimal T on the day 0 and maximal T on the day ÿ 1. D T2: difference between maximal T on the day 0 and minimal T on the day ÿ 1.
b c
Based on the number of days elapsed between two successive SP-admissions, all patients were classified in groups depending on whether the aforementioned interval was up to 2, 3, or 4 days. In the analyzed period 352 (53.41%) patients were admitted with the interval between two successive admissions not longer than 2 days; 3- and 4-day intervals were associated with 454 (68.89%) and 503 (76.32%) patients, respectively (Fig. 2).
DISCUSSION This study did not confirm significant correlation of the occurrence of SP with either particular months or with seasons of the year. Even in studies confirming the more frequent occurrence of SP in winter months, these differences are not so clear; in the study by Accard et al. (6), the frequency of SP was highest from October to March, but the maximal frequency was however registered between November and February. Likewise, other studies confirming the increased occurrence of SP in the winter, simultaneously
AEP Vol. 15, No. 3 March 2005: 185–190
confirm the increasing frequency of SP in another period of the year (7). In about 70% of SP-days AP was stable, whilst AP was falling and rising in 15.1% and 16.6% SP-days, respectively. Accard (6) found similar distribution in his study, but without taking into account AP trend in non-SP days; in our study, no significant difference concerning AP trend (stable, falling, or rising) was found between SP and non-SP days. Concerning the influence of AP changes, it is likely that repeated changes in volume of the air-containing lung cysts can weaken their wall, causing their rupture and the onset of SP. A widespread explanation is that the principal contributive factor could be a fall of AP of at least 10 mBar in a 24-hour period (8). Some authors state that abrupt AP changes, independent of their direction, are more important for the onset of SP than simple falls of AP by itself (9). Vesicles with valves, representing residual lesions in the lungs, typical in localized emphysema independent of age, are also reported as causes of lung cysts’ ‘‘isolation’’ (10). Based on rare studies of SP in the acute phase, closure of small airways occurs in the affected lung, causing impairment of the distribution of ventilation (11). One should expect that the influence of significant AP changes will be more efficient if the period of exposure is longer (for instance, 3 days, instead of 24 hours, preceding SP). According to one of the most detailed studies of this problem by Scott et al. (12), only one-half of SP-patients were exposed to significant AP-changes during the 4 days preceding SP. In our study, only 25% of patients were exposed to 1 to 4 such episodes, without significant differences in the exposure pattern between patients with primary SP and SP associated with COPD. This finding does
FIGURE 1. Distribution of weather phases in days with pneumothorax. SP Days: days with pneumothorax.
AEP Vol. 15, No. 3 March 2005: 185–190
FIGURE 2. Admission pattern of patients with spontaneous pneumothorax. <2 days, <3 days, <4 days: interval between two successive admissions <2, 3, or 4 days respectively; percent values correspond to presented intervals.
not mean that abrupt AP changes do not influence the onset of SP, but indicates the possible role of other factors whose influence may not be evident in the analysis of this type. Our finding that the observed frequency of exposure to 3 or more than 4 episodes of significant AP-changes exceeds the expected frequency supports such a statement. Likewise, if AP-changes are analyzed in lesser intervals, nearly 30% of patients were exposed to significant AP changes. Hence, the lesser the interval of measurement, the more patients exposed to such changes. Most studies analyzing the influence of atmospheric studies on SP do not consider T changes. Our study did not find significant correlation between T-changes and SP occurrence, but there is some evidence indicating that days with SP are preceded by significant T rises (13). Our study confirmed the well known pattern of SP-onset in series, or ‘‘clusters.’’ In our study, 53.41% of patients were admitted to a hospital with an interval between the two successive admissions not exceeding 2 days, whilst in more than 75% of patients that interval was longer than 4 days. In Smith’s study (13), admission of 61% and 83% of patients followed a pattern of 2-day and 4-day clusters, respectively. These results, however, do not confirm whether atmospheric factors are dominant contributive factors in SP occurrence. Regarding the relationship between weather phases and SP, to our knowledge, there are no published studies dealing with this problem. According to a similar study, in which the influence of weather phases on asthma exacerbation was analyzed, weather phases 2ts and 5hv (unfavorable for SP in our study), together with a few other phases, were also found to be unfavorable for this serious condition (14). It is an expected finding, keeping in mind the fact that asthma and SP share some common pathophysiological mechanisms (inflammation of small airways) that are considered to
Bulajich et al. ATMOSPHERIC FACTORS AND PNEUMOTHORAX
189
participate in the aforementioned ‘‘isolation’’ of aircontaining cysts in the lungs of SP-patients (15), (16). The main features of weather phases 2ts (anticyclonic situation with warm and dry weather) and 5hv (passing of the cold front), that significantly correlated with SP, are fast changes of several meteorological factors, requiring from the organism prompt adaptation to the new situation. Does this significant correlation have some biological sense and through which mechanisms? In fact, none of pathophysiological mechanisms is directly related to blebs/ bullae rupture, but only to factors leading to their ‘‘isolation’’ from the surrounding lung tissue, as already described. It was demonstrated that stimuli from environmental factors influence our immune system through adrenergic receptors on granulocytes and cholinergic receptors on lymphocytes; activation of granulocytes induces infection/inflammation or tissue destruction, whereas activation of lymphocytes induces allergic disease (17). As already mentioned, inflammation of small airways is supposed to be the main cause of blebs isolation. Some reports suggest that bronchoconstriction induced by moistening of air in the airways might play a role in the occurrence of SP (18). Some recent studies suggest that some fluid imbalance in the small airways, together with pollutant exposure, could cause airway obstruction with segmental increase in airway resistance and rise in distal pressure (19, 20). In an urban environment with high pollution, as in our study, the concentration of heavy (positive) is higher than the concentration of light (negative) ions. The fusion of heavy ions and pollutant particles causes local irritation, airway inflammation, and increased epithelial permeability (21). It is also well known that the influence of ions coincides (or even precedes) the incoming front. This directly supports our finding of significant correlation of SP-onset and passing of cold front (phase 5hv), being a carrier of positive ions. In the analyzed period, increasing number of SP admissions was registered during changes of weather phases with anticyclonic activity (2hs, 2ts, 2tv) to phases representing fast passage of fronts (5hv). These phases are extremely unstable, with quickly changing meteorological situation, leading to the change of the electrical potential of the atmosphere, requiring from the organism prompt reaction and adaptation. These facts could help explain the relationship between meteorological factors and host response mechanisms leading to the onset of SP. With regard to acute respiratory infections and asthma, the two conditions potentially sharing similar mechanisms of inflammation with SP, one earlier analysis in the same geographic location as in our study (14), revealed situations with fog (cold, wet weather [6hv]) as unfavorable for both acute respiratory infections and asthma; moreover, anticyclone with warm and dry weather (phase 3ats), very similar to phase 2ts, which in our study was significantly related to
190
Bulajich et al. ATMOSPHERIC FACTORS AND PNEUMOTHORAX
AEP Vol. 15, No. 3 March 2005: 185–190
TABLE 5. Results of studies reporting on the influence of atmospheric pressure (AP), temperature (T), and weather phases (w.ph) on the occurrence of SP Author (ref) Accard (6) Schmerber (25)1 Jensen (26)2 Bense (8)3 Scott (12)4 Smith (13)5 Suarez-Varela (20) Bulajich (2001.)
4. Lowry RH, Wood AM, Higenbottam TW. The effect of pH and osmolarity on aerosol induced cough in normal volunteers. Clin Sci. 1988;74:373– 376.
No. of patients
AP
T
w. ph
95 116 180 282 685 115 69 659
NS NS NS S HS NS NS NS
NS NA NA NA NA S NA NS
NA NA NA NA NA NA NA S
NS, not significant; S, significant; HS, highly significant; NA, not analyzed. 1 Single falls of AP of O 10 mBar in a 3-day period. 2 Incidence of SP in hours of DAP below or above the yearly mean, DAP immediately prior to SP not analyzed. 3 Admission rate significantly increased 48 hours after a fall in AP of at least 10 mBar in 24 hours. 4 Observed significantly greater than expected number of exposures to unusual DAP (> 10 mBar). 5 Temperature rise of 0.57 C from the day prior to SP vs. 0.08 C fall on the days without SP.
SP-onset, was also an unfavorable situation for asthma attacks. The explanation of these findings is more complex. Although it was confirmed that frequency of asthma exacerbations was higher on misty or foggy nights than on clear nights, it is not clear whether this frequency is related to airborne water droplets, or to the meteorological condition that causes mist or fog (22). Because a higher atmospheric temperature on misty or foggy nights indicates a larger saturated amount of airborne water droplets, other recent studies suggest that mist and fog may be a stimulus for bronchoconstriction (23, 24). In brief, expressing different atmospheric factors in the form of weather phases facilitates the explanation of the biological sense of the obtained results, because the majority of studies analyzing the influence of isolated atmospheric factors on SP onset did not confirm significant causative relationship (Table 5). CONCLUSION Among all analyzed meteorological factors, this study revealed only weather phases 2ts and 5hv as unfavorable in terms of the occurrence of SP. However, other possible external contributive factors must be investigated and further research is necessary to identify precise mechanisms of the biological response. REFERENCES 1. Ziser A, Vaandnen A, Melamed Y. Diving and chronic spontaneous pneumothorax. Chest. 1985;87:264–265.
5. Laitinen A. Autonomic innervation of the human respiratory tract as revealed by histochemical method and ultrastructural method. Eur J Respir Dis. 1985;66(suppl. 140):1–42. 6. Accard JL, Patte F, Combes F, Sorin C, Gaillard F, Garel L, et al. Pneumothorax spontane. Rev Tuberc Pneumonol. 1972;36:431–446. 7. Macquet V, Lafitte P, Rogeaux Y, Pinchon R. Influence des conditions meteorologiques dans la survenue du pneumothorax spontane. Lille Medical. 1974;19:901–906. 8. Bense I. Spontaneous pneumothorax related to falls in atmospheric pressure. Eur J Respir Dis. 1984;65:544–546. 9. Ohata M, Suzuki H. Pathogenesis of spontaneous pneumothorax. With special reference to the ultrastructure of emphysematous bullae. Chest. 1980;77:771–776. 10. Perry MA. On spontaneous pneumothorax. Q J Med. 1939;8:1. 11. Anthonisen NR. Regional lung function in spontaneous pneumothorax. Am Rev Respir Dis. 1977;115:873–876. 12. Scott GC, Berger R, McKean. The role atmospheric pressure variation in the development of Spontaneous pneumothorax. Am Rev Respir Dis. 1989;139:659–662. 13. Smit H, Deville W, Schramel M, Schreurs A, Sutedja T, Postmus P. Atmospheric pressure changes and outdoor temperature changes in relation to spontaneous pneumothorax. Chest. 1999;116:676–681. 14. Djekich M, Cvetanovich S, Stozinic S. Bronchial asthma. In: Djekic M, ed. Influence of Weather on Human Health. Belgrade: Naucna knjiga; 1981:82–90 15. Laitinen LA, Heino M, Laitinen A, Kava T, Haahtela T. Damage of the airway epithelium and bronchial reactivity in patients with asthma. Am Rev Respir Dis. 1985;131:559–606. 16. McFadden ER, Ingram RH. Pathogenesis of asthma. J Allergy Clin Immunol. 1984;73:413–428. 17. Abo T, Kawamura T. [Immune system and diseases that vary with circumstance and physical condition]. Rinsho Byori. 1997;45:3–12. 18. Ozenne G, Poignie P, Lemercier JP, Nouvet G, Grancher G. Meteorological conditions and spontaneous pneumothorax. Retrospective study of 165 cases in the Rouen area. Rev Pneumol Clin. 1984;40:27–33. 19. Sok M, Mikulecky M, Erzen J. Onset of spontaneous pneumothorax and synodic lunar cycle. Med Hypotheses. 2001;57:638–641. 20. Morales Suarez-Varela S, Martinez-Selva MI, Llopis-Gonzales A, Martinez-Jimeno JL, Plaza-Valia P. Spontaneous pneumothorax related with climatic characteristics in the Valencia area (Spain). Eur J Epidemiol. 2000;16:193–198. 21. D’Amato G, Liccardi G, D’Amato M. Environmental risk factors (outdoor air pollution and climatic changes) and increased trend of respiratory allergy. J Investig Allergol Clin Immunol. 2000;10:123–128. 22. Kashiwabara K, Kohrogi H, Ota K, Moroi T. High frequency of emergency room visits of asthmatic children on misty or foggy nights. J Asthma. 2002; 39:711–717. 23. Kashiwabara K, Itonaga K, Moroi T. Airborne water droplets in mist or fog may affect nocturnal attacks in asthmatic children. J Asthma. 2003; 40:405–411. 24. Schappi GF, Taylor PE, Pain MC, Cameron PA, Dent AW, Staff IA, et al. Concentrations of major grass group 5 allergens in pollen grains and atmospheric particles: Implications for hay fever and allergic asthma sufferers sensitized to grass pollen allergens. Clin Exp Allergy. 1999; 29:633–641.
2. Sadikot RT, Greene T, Meadows KK. Recurrence of primary spontaneous pneumothorax. Thorax. 1997;52:805–809.
25. Schmerber J, Manderlier T. Effect of atmospheric pressure on the occurence of spontaneous pneumothorax. Acta tuberculosea et pneumologica belgica. 1978;69:3–4.
3. Fry WA, Paape K. Pneumothorax. In: Shelds TW, ed. General Thoracic Surgery. Philadelphia: Lea and Febiger; 2000;675–686.
26. Jensen BS. Pneumothorax simplex and atmospheric pressure. Ugeskrift for Laeger. 1980:1603–1606.
PURPOSE: To analyze the influence of meteorological factors such as atmospheric pressure (AP), outdoor temperature (T) changes, and weather phases (WP) on the occurrence of spontaneous pneumothorax (SP). METHODS: Retrospective study, including 659 patients with primary SP and SP associated with chronic obstructive pulmonary disease (COPD), conservatively treated in a 5-year period. In the analyzed period, 548 days with SP were compared both with 3 days preceding the onset of SP and with days without pneumothorax. The comparison was made depending on weather phases and on different aspects of AP and T. RESULTS: Seasons of the year did not significantly influence the occurrence of SP. No significant difference was found between SP and non-SP days depending on different aspects of AP and T changes. Most patients were admitted in ‘‘clusters’’ with not more than 2, 3, or 4 days between two successive admissions. The occurrence of SP was significantly correlated with weather phases 2ts (anticyclonic situation with warm and dry weather) and 5 hv (passing of the cold front). Biological sense and possible explanations of this correlation are discussed. CONCLUSIONS: Among all analyzed meteorological factors, significant correlation was found only between weather phases 2ts and 5hv and the occurrence of SP. Ann Epidemiol 2005;15:185–190. Ó 2004 Elsevier Inc. All rights reserved. KEY WORDS:
Weather Phases, Pneumothorax, Atmospheric Pressure, Temperature.
INTRODUCTION The explanation of the influence of meteorological factors to the onset of spontaneous pneumothorax (SP) is the assumption that as long as free communication exists between subpleural air-containing cysts and atmospheric air, pneumothorax will not ensue, even in presence of major atmospheric pressure (AP) changes. Once this communication is disrupted, these cysts become isolated from the surrounding lung tissue, changing their morphology according to Boyle-Marriott’s law (P ! V Z const.), i.e., a fall in AP causes an increase in their volume (1). Although the morphology and ultrastructure of causative lesions in primary SP are well known, the reason for rupture of air-containing cysts is not absolutely clear (2). Broad
From the Intensive Care Unit (B.B.), and Clinic for Thoracic Surgery (D.S., D.M.), Institute for Lung Diseases, Clinical Center of Serbia, Belgrade, Serbia and Montenegro; Republic Hydrometeorological Service of Serbia, Belgrade, Serbia and Montenegro (R.V.K.); and Institute for Medical Statistics, Faculty of Medicine, Belgrade, Serbia and Montenegro (M.G.). Address correspondence to: Dragan Subotich, M.D., Ph.D., F.C.C.P., Clinic for Thoracic Surgery, Institute for Lung Diseases, Clinical Center of Serbia, Visegradska 26/20, Belgrade 11000, Serbia and Montenegro. Tel.: C381-11-361-5559; Fax: C381-11-646-988. E-mail: [email protected]; [email protected] Received December 15, 2003; accepted April 25, 2004. Ó 2004 Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010
consensus concerning the role of atmospheric factors in SP does not exist (3). The aim of this study, apart from the analysis of the influence of AP and outdoor temperature changes on the occurrence of SP, was to assess the relationship between SP and different weather phases in days with SP and during 3 preceding days. The term ‘‘weather phase’’ summarizes the influence of several meteorological factors, indicates the dynamics of their changes, thus directly correlates them with possible host response mechanisms (4, 5).
METHODS AND MATERIALS Patients The study included 659 patients with SP conservatively treated at the Institute for Lung Diseases in Belgrade between January 1990 and December 1994. Only patients with primary SP and SP associated with chronic obstructive pulmonary disease (COPD) were included. Of a total number of 659 admitted patients, 565 (85.7%) patients had primary SP and 94 (14.3%) patients had SP associated with COPD. There were 529 (80.3%) males and 130 (19.7%) females. Three hundred fifty (53.11%) patients belonged to the age group 21 to 40 years, whilst only 76 (11.54%) and 77 (11.68%) patients were younger than 20 years and older than 60 years, respectively. In 59.33%, SP was right sided. 1047-2797/05/$–see front matter doi:10.1016/j.annepidem.2004.04.006
186
Bulajich et al. ATMOSPHERIC FACTORS AND PNEUMOTHORAX
The criterion for the onset of SP was the day of the onset of symptoms reported by the patient. Only patients with the exact date of the onset of SP were included in the study. Also, only patients who contracted SP in the Belgrade region were included in the study, provided that they were there at least one day before the onset of symptoms. Meteorological Data The Republic Hydrometeorological Service in Belgrade provided atmospheric pressure (AP) and temperature (T) measurements. For each of the 1825 days in the analyzed 5year period, AP (mBar) measured at 7 hours, 14 hours, and 21 hours, as well as mean, maximal, and minimal AP values were registered. Temperature ( C) was measured at the same intervals, together with mean, maximal, and minimal T values. The aforementioned data were registered on the day of the onset of SP (‘‘day 0’’) and during 3 previous days (‘‘day ÿ 1,’’ ‘‘day ÿ 2,’’ and ‘‘day ÿ 3’’). Each day in the analyzed period corresponded to one of 20 weather phases (according to Hess-Berezowsky) that were registered and entered in the database. Names and main characteristics of weather phases analyzed in this study are as follows: 1hsdmild influence of anticyclone with cold and dry weather; 1tsdmild influence of anticyclone with warm and dry weather; 2hsdanticyclonic situation with cold and dry weather; 2tsdanticyclonic situation with warm and dry weather; 2tvdanticyclonic situation with warm and wet weather; 3atsdcenter of the anticyclone, warm and dry weather; 3ftsdcenter of the cyclone, warm and dry weather; 4hvdfront side of the cyclone with cold and wet weather; 4tsdwarm, dry weather, preceding the weather change; 4tvdfront side of the cyclone with warm and wet weather; 5 hs, 5 hv, 5 ts, 5tvdpassing of fronts (5hs: cold front; 5hv: occlusion, cold front type; 5ts: warm front; 5tv: occlusion, warm front type); 6 zhsdweather stabilization with cold and dry weather; 6zhvdweather stabilization with cold and wet weather; 6ztsddry weather between two fronts; 6ztvdwarm and wet weather preceding the cold front; 6hsdweather situation with temperature inversion; and 6hvdfog during the whole day. Data Analysis In the analyzed period there were 548 days with SP. These days were compared both with 3 days preceding the onset of SP and with all days without pneumothorax in the analyzed period. The comparison was made depending on different meteorological factorsdmean, maximal, and minimal AP and T values; maximal daily AP- and T-amplitudes; significant AP variations (AP amplitudes of at least 10 mBar in a 24-hour interval). For this type of the analysis, expected frequency of significant AP changes was calculated and compared with registered frequency. Days with
AEP Vol. 15, No. 3 March 2005: 185–190
pneumothorax were also compared with 3 preceding days depending on weather phases. Comparisons depending on different meteorological factors were made both with type of SP independently and depending on whether SP was primary or associated with COPD. Statistical analysis included distribution of frequencies, c2 test, t-test for paired samples, t-test for unpaired samples, and Wilcoxon-Mann-Whitney U-test.
RESULTS Seasons and the Occurrence of SP The frequency of SP occurrence was nearly equal in all seasons of the year. In spring, SP occurred in 171 (25.9%) and in summer in 163 (24.7%) patients; in most patients (174 [26.4%]) SP developed in autumn, whilst winter was the season with minimal (151 [23%]) number of patients with SP. The monthly distribution of admitted SP patients was different depending on the year. The most prominent variations were registered in February and Decemberdthe number of admitted patients in February 1992 was threefold the number for the same month in 1994; similar difference exists between December 1994 and December 1992. Less prominent variations were registered in April, June, and October. General Meteorological Features of SP-days and Non-SP Days On SP-days AP varied between 980.10 and 1023.50 mBar (mean value, 1002.19 G 7.21). Neither for AP nor for T were differences between SP-days and non-SP days statistically significant. Maximal daily AP amplitudes on SP-days were between 0.10 and 18.80 mBar (mean value, 2.89 G 2.32); on non-SP days, maximal amplitudes were between 0 and 16.10 mBar (mean value, 2.90 G 2.35). On SP-days, maximal daily T amplitudes were between ÿ 1.60 C and 19.80 C, mean value 9.46 G 3.98. Neither for AP nor for T were differences between SP-days and non-SP days statistically significant. Distribution of different AP intervals on SP-days and on non-SP days is presented in Table 1. Differences between SP-days and non SP-days depending on these intervals are not statistically significant. Similarly, differences between each group and the distribution in the entire analyzed period are not statistically significant. The analysis of the AP trend (stable, rising, or falling) on days with and without SP is also presented in Table 1. With AP change of at least 10 mBar accepted as significant, in 93.06% of days with SP there were no significant fluctuations in relation to the preceding day. Differences between SP-days and non-SP days are not statistically significant. With AP changes of at least 5 mBar as significant, although
AEP Vol. 15, No. 3 March 2005: 185–190
Bulajich et al. ATMOSPHERIC FACTORS AND PNEUMOTHORAX
TABLE 1. Atmospheric pressure intervals and trend on days with and without pneumothorax Days without SP
Days with SP
AP interval (mBar)
N
%
N
%
! 980 980–1000 1001–1010 1011–1020 1021–1030
102 522 470 154 29
7.98 40.87 36.80 12.05 2.30
47 214 221 61 5
8.57 39.05 40.32 11.13 0.93
1277
100
548
100
Total
Days without SP AP trend* Falling Stable Rising Total
Total N
149 8.16 736 40.33 691 37.86 215 11.78 34 1.87 1825
Days with SP
n
%
n
%
46 1206 25 1277
3.60 94.22 2.18 100
18 510 20 548
3.30 93.06 3.64 100
%
100
Total n
%
64 3.50 1716 94.02 45 2.48 1825 100
P O 0.05
187
significant AP-changes of patients with primary SP is identical to the exposure of SP-patients with COPD. The expected frequency of significant AP changes was calculated and compared with the observed frequency, both for the entire group and separately for primary SP and SP with COPD (Table 3). For the entire group, three or more than four exposures to such AP changes during 24 hours were significantly more frequent than expected. Outdoor Temperature Temperature differences between SP-days and non-SP days compared with the preceding day for both groups are presented in Table 4. Although in days preceding the occurrence of SP a certain temperature drop was registered (compared with non-SP days that were preceded by a day with a light temperature rise), no significant differences in mean temperature between SP-days vs. the preceding days were found.
*AP changes related to the preceding day (7-hour day – 1:7-hour day 0).
Weather Phases AP was stable in only 68.24% of SP-days (data not shown), differences between SP-days and non-SP days were not statistically significant.
Atmospheric Pressure Significant AP changes (DAP of at least 10 mBar in a 24hour interval) during the 3 days preceding SP are presented in Table 2. Of a total of 659 patients with SP, in 491 (74.65%) patients no significant AP changes were registered; the rest (?25%) of the patients were exposed to at least one, maximally to six, significant AP changes. Most of them (140 [16.23%]) were exposed to between one and three such episodes; 15 (2.30%) patients were exposed to four significant episodes, whilst five and six episodes were registered in 6 (0.91%) patients each. The exposure to TABLE 2. Significant AP changes (G 10 mBar) during three days preceding SP (24-hour interval)
No. of changes 0 1 2 3 4 5 6 Total
Primary SP
SP C COPD
n
%
n
%
n
%
420 35 46 38 14 6 6 565
74.30 6.20 8.20 6.70 2.50 1.10 1.10 100
72 3 8 10 1 0 0 94
76.60 3.20 8.50 10.60 1.10 0 0 100
492 38 54 48 15 6 6 659
74.60 5.80 8.20 7.30 2.30 0.90 0.90 100
P O 0.05
Total
In the analyzed period, SP occurred most frequently in presence of two weather phasesdphase 2ts and phase 5hv, each with 120 (21.9%) SP days (Fig. 1). The third most frequent was the weather phase 2hs with 11.9% SP days. Weather phase 6zhv was associated with 8.8% SP days. No differences in dominant weather phases distribution existed during the 3 days preceding SP. Concerning dominant combinations of weather phases, during the 2 days preceding SP, the most frequent pairs of weather phases were 2ts-2ts and 6zhv-5hv registered in 11 and 10 patients respectively. Neither of analyzed combinations was of statistical significance. Occurrence of Spontaneous Pneumothorax in Series (‘‘Clusters’’) In the analyzed 5-year period there were 548 (29.96%) days with SP. Most days (456 [83.21%]) had only one SP admission; on 75 (13.68%) and 16 (2.91%) days, two and three SP admissions were registered, respectively. Four SP admissions were registered on only one day during the entire period. TABLE 3. Comparison of the observed with the expected frequency of significant AP changes (24 hour interval) Number of exposures to significant AP changes (Os/Oc) 0 Primary SP 419/418 SP C COPD 72/70 All 491/488
1
2
35/51 3/9 38/60
46/45 8/7 54/52
Os/oc = observed frequency/expected frequency. *Statistically significant difference.
3 38/31 10/5 48/36*
4
O4
14/11 1/2 15/13
12/8 0/1 12/9*
188
Bulajich et al. ATMOSPHERIC FACTORS AND PNEUMOTHORAX
TABLE 4. Differences in the outdoor temperature between the ‘‘day 0’’ and the ‘‘day ÿ 1’’ on days with and without SP D T in relation to day ÿ 1 a
Mean D T1b D T2c
Days with SP
Days without SP
ÿ0.02 G 2.82 ÿ9.63 G 3.65 9.44 G 4.92
0.01 G 2.86 ÿ9.30 G 3.97 9.39 G 4.77
P O 0.05 a
Mean: difference in mean T between the day ÿ 1 and the day 0. D T1: difference between minimal T on the day 0 and maximal T on the day ÿ 1. D T2: difference between maximal T on the day 0 and minimal T on the day ÿ 1.
b c
Based on the number of days elapsed between two successive SP-admissions, all patients were classified in groups depending on whether the aforementioned interval was up to 2, 3, or 4 days. In the analyzed period 352 (53.41%) patients were admitted with the interval between two successive admissions not longer than 2 days; 3- and 4-day intervals were associated with 454 (68.89%) and 503 (76.32%) patients, respectively (Fig. 2).
DISCUSSION This study did not confirm significant correlation of the occurrence of SP with either particular months or with seasons of the year. Even in studies confirming the more frequent occurrence of SP in winter months, these differences are not so clear; in the study by Accard et al. (6), the frequency of SP was highest from October to March, but the maximal frequency was however registered between November and February. Likewise, other studies confirming the increased occurrence of SP in the winter, simultaneously
AEP Vol. 15, No. 3 March 2005: 185–190
confirm the increasing frequency of SP in another period of the year (7). In about 70% of SP-days AP was stable, whilst AP was falling and rising in 15.1% and 16.6% SP-days, respectively. Accard (6) found similar distribution in his study, but without taking into account AP trend in non-SP days; in our study, no significant difference concerning AP trend (stable, falling, or rising) was found between SP and non-SP days. Concerning the influence of AP changes, it is likely that repeated changes in volume of the air-containing lung cysts can weaken their wall, causing their rupture and the onset of SP. A widespread explanation is that the principal contributive factor could be a fall of AP of at least 10 mBar in a 24-hour period (8). Some authors state that abrupt AP changes, independent of their direction, are more important for the onset of SP than simple falls of AP by itself (9). Vesicles with valves, representing residual lesions in the lungs, typical in localized emphysema independent of age, are also reported as causes of lung cysts’ ‘‘isolation’’ (10). Based on rare studies of SP in the acute phase, closure of small airways occurs in the affected lung, causing impairment of the distribution of ventilation (11). One should expect that the influence of significant AP changes will be more efficient if the period of exposure is longer (for instance, 3 days, instead of 24 hours, preceding SP). According to one of the most detailed studies of this problem by Scott et al. (12), only one-half of SP-patients were exposed to significant AP-changes during the 4 days preceding SP. In our study, only 25% of patients were exposed to 1 to 4 such episodes, without significant differences in the exposure pattern between patients with primary SP and SP associated with COPD. This finding does
FIGURE 1. Distribution of weather phases in days with pneumothorax. SP Days: days with pneumothorax.
AEP Vol. 15, No. 3 March 2005: 185–190
FIGURE 2. Admission pattern of patients with spontaneous pneumothorax. <2 days, <3 days, <4 days: interval between two successive admissions <2, 3, or 4 days respectively; percent values correspond to presented intervals.
not mean that abrupt AP changes do not influence the onset of SP, but indicates the possible role of other factors whose influence may not be evident in the analysis of this type. Our finding that the observed frequency of exposure to 3 or more than 4 episodes of significant AP-changes exceeds the expected frequency supports such a statement. Likewise, if AP-changes are analyzed in lesser intervals, nearly 30% of patients were exposed to significant AP changes. Hence, the lesser the interval of measurement, the more patients exposed to such changes. Most studies analyzing the influence of atmospheric studies on SP do not consider T changes. Our study did not find significant correlation between T-changes and SP occurrence, but there is some evidence indicating that days with SP are preceded by significant T rises (13). Our study confirmed the well known pattern of SP-onset in series, or ‘‘clusters.’’ In our study, 53.41% of patients were admitted to a hospital with an interval between the two successive admissions not exceeding 2 days, whilst in more than 75% of patients that interval was longer than 4 days. In Smith’s study (13), admission of 61% and 83% of patients followed a pattern of 2-day and 4-day clusters, respectively. These results, however, do not confirm whether atmospheric factors are dominant contributive factors in SP occurrence. Regarding the relationship between weather phases and SP, to our knowledge, there are no published studies dealing with this problem. According to a similar study, in which the influence of weather phases on asthma exacerbation was analyzed, weather phases 2ts and 5hv (unfavorable for SP in our study), together with a few other phases, were also found to be unfavorable for this serious condition (14). It is an expected finding, keeping in mind the fact that asthma and SP share some common pathophysiological mechanisms (inflammation of small airways) that are considered to
Bulajich et al. ATMOSPHERIC FACTORS AND PNEUMOTHORAX
189
participate in the aforementioned ‘‘isolation’’ of aircontaining cysts in the lungs of SP-patients (15), (16). The main features of weather phases 2ts (anticyclonic situation with warm and dry weather) and 5hv (passing of the cold front), that significantly correlated with SP, are fast changes of several meteorological factors, requiring from the organism prompt adaptation to the new situation. Does this significant correlation have some biological sense and through which mechanisms? In fact, none of pathophysiological mechanisms is directly related to blebs/ bullae rupture, but only to factors leading to their ‘‘isolation’’ from the surrounding lung tissue, as already described. It was demonstrated that stimuli from environmental factors influence our immune system through adrenergic receptors on granulocytes and cholinergic receptors on lymphocytes; activation of granulocytes induces infection/inflammation or tissue destruction, whereas activation of lymphocytes induces allergic disease (17). As already mentioned, inflammation of small airways is supposed to be the main cause of blebs isolation. Some reports suggest that bronchoconstriction induced by moistening of air in the airways might play a role in the occurrence of SP (18). Some recent studies suggest that some fluid imbalance in the small airways, together with pollutant exposure, could cause airway obstruction with segmental increase in airway resistance and rise in distal pressure (19, 20). In an urban environment with high pollution, as in our study, the concentration of heavy (positive) is higher than the concentration of light (negative) ions. The fusion of heavy ions and pollutant particles causes local irritation, airway inflammation, and increased epithelial permeability (21). It is also well known that the influence of ions coincides (or even precedes) the incoming front. This directly supports our finding of significant correlation of SP-onset and passing of cold front (phase 5hv), being a carrier of positive ions. In the analyzed period, increasing number of SP admissions was registered during changes of weather phases with anticyclonic activity (2hs, 2ts, 2tv) to phases representing fast passage of fronts (5hv). These phases are extremely unstable, with quickly changing meteorological situation, leading to the change of the electrical potential of the atmosphere, requiring from the organism prompt reaction and adaptation. These facts could help explain the relationship between meteorological factors and host response mechanisms leading to the onset of SP. With regard to acute respiratory infections and asthma, the two conditions potentially sharing similar mechanisms of inflammation with SP, one earlier analysis in the same geographic location as in our study (14), revealed situations with fog (cold, wet weather [6hv]) as unfavorable for both acute respiratory infections and asthma; moreover, anticyclone with warm and dry weather (phase 3ats), very similar to phase 2ts, which in our study was significantly related to
190
Bulajich et al. ATMOSPHERIC FACTORS AND PNEUMOTHORAX
AEP Vol. 15, No. 3 March 2005: 185–190
TABLE 5. Results of studies reporting on the influence of atmospheric pressure (AP), temperature (T), and weather phases (w.ph) on the occurrence of SP Author (ref) Accard (6) Schmerber (25)1 Jensen (26)2 Bense (8)3 Scott (12)4 Smith (13)5 Suarez-Varela (20) Bulajich (2001.)
4. Lowry RH, Wood AM, Higenbottam TW. The effect of pH and osmolarity on aerosol induced cough in normal volunteers. Clin Sci. 1988;74:373– 376.
No. of patients
AP
T
w. ph
95 116 180 282 685 115 69 659
NS NS NS S HS NS NS NS
NS NA NA NA NA S NA NS
NA NA NA NA NA NA NA S
NS, not significant; S, significant; HS, highly significant; NA, not analyzed. 1 Single falls of AP of O 10 mBar in a 3-day period. 2 Incidence of SP in hours of DAP below or above the yearly mean, DAP immediately prior to SP not analyzed. 3 Admission rate significantly increased 48 hours after a fall in AP of at least 10 mBar in 24 hours. 4 Observed significantly greater than expected number of exposures to unusual DAP (> 10 mBar). 5 Temperature rise of 0.57 C from the day prior to SP vs. 0.08 C fall on the days without SP.
SP-onset, was also an unfavorable situation for asthma attacks. The explanation of these findings is more complex. Although it was confirmed that frequency of asthma exacerbations was higher on misty or foggy nights than on clear nights, it is not clear whether this frequency is related to airborne water droplets, or to the meteorological condition that causes mist or fog (22). Because a higher atmospheric temperature on misty or foggy nights indicates a larger saturated amount of airborne water droplets, other recent studies suggest that mist and fog may be a stimulus for bronchoconstriction (23, 24). In brief, expressing different atmospheric factors in the form of weather phases facilitates the explanation of the biological sense of the obtained results, because the majority of studies analyzing the influence of isolated atmospheric factors on SP onset did not confirm significant causative relationship (Table 5). CONCLUSION Among all analyzed meteorological factors, this study revealed only weather phases 2ts and 5hv as unfavorable in terms of the occurrence of SP. However, other possible external contributive factors must be investigated and further research is necessary to identify precise mechanisms of the biological response. REFERENCES 1. Ziser A, Vaandnen A, Melamed Y. Diving and chronic spontaneous pneumothorax. Chest. 1985;87:264–265.
5. Laitinen A. Autonomic innervation of the human respiratory tract as revealed by histochemical method and ultrastructural method. Eur J Respir Dis. 1985;66(suppl. 140):1–42. 6. Accard JL, Patte F, Combes F, Sorin C, Gaillard F, Garel L, et al. Pneumothorax spontane. Rev Tuberc Pneumonol. 1972;36:431–446. 7. Macquet V, Lafitte P, Rogeaux Y, Pinchon R. Influence des conditions meteorologiques dans la survenue du pneumothorax spontane. Lille Medical. 1974;19:901–906. 8. Bense I. Spontaneous pneumothorax related to falls in atmospheric pressure. Eur J Respir Dis. 1984;65:544–546. 9. Ohata M, Suzuki H. Pathogenesis of spontaneous pneumothorax. With special reference to the ultrastructure of emphysematous bullae. Chest. 1980;77:771–776. 10. Perry MA. On spontaneous pneumothorax. Q J Med. 1939;8:1. 11. Anthonisen NR. Regional lung function in spontaneous pneumothorax. Am Rev Respir Dis. 1977;115:873–876. 12. Scott GC, Berger R, McKean. The role atmospheric pressure variation in the development of Spontaneous pneumothorax. Am Rev Respir Dis. 1989;139:659–662. 13. Smit H, Deville W, Schramel M, Schreurs A, Sutedja T, Postmus P. Atmospheric pressure changes and outdoor temperature changes in relation to spontaneous pneumothorax. Chest. 1999;116:676–681. 14. Djekich M, Cvetanovich S, Stozinic S. Bronchial asthma. In: Djekic M, ed. Influence of Weather on Human Health. Belgrade: Naucna knjiga; 1981:82–90 15. Laitinen LA, Heino M, Laitinen A, Kava T, Haahtela T. Damage of the airway epithelium and bronchial reactivity in patients with asthma. Am Rev Respir Dis. 1985;131:559–606. 16. McFadden ER, Ingram RH. Pathogenesis of asthma. J Allergy Clin Immunol. 1984;73:413–428. 17. Abo T, Kawamura T. [Immune system and diseases that vary with circumstance and physical condition]. Rinsho Byori. 1997;45:3–12. 18. Ozenne G, Poignie P, Lemercier JP, Nouvet G, Grancher G. Meteorological conditions and spontaneous pneumothorax. Retrospective study of 165 cases in the Rouen area. Rev Pneumol Clin. 1984;40:27–33. 19. Sok M, Mikulecky M, Erzen J. Onset of spontaneous pneumothorax and synodic lunar cycle. Med Hypotheses. 2001;57:638–641. 20. Morales Suarez-Varela S, Martinez-Selva MI, Llopis-Gonzales A, Martinez-Jimeno JL, Plaza-Valia P. Spontaneous pneumothorax related with climatic characteristics in the Valencia area (Spain). Eur J Epidemiol. 2000;16:193–198. 21. D’Amato G, Liccardi G, D’Amato M. Environmental risk factors (outdoor air pollution and climatic changes) and increased trend of respiratory allergy. J Investig Allergol Clin Immunol. 2000;10:123–128. 22. Kashiwabara K, Kohrogi H, Ota K, Moroi T. High frequency of emergency room visits of asthmatic children on misty or foggy nights. J Asthma. 2002; 39:711–717. 23. Kashiwabara K, Itonaga K, Moroi T. Airborne water droplets in mist or fog may affect nocturnal attacks in asthmatic children. J Asthma. 2003; 40:405–411. 24. Schappi GF, Taylor PE, Pain MC, Cameron PA, Dent AW, Staff IA, et al. Concentrations of major grass group 5 allergens in pollen grains and atmospheric particles: Implications for hay fever and allergic asthma sufferers sensitized to grass pollen allergens. Clin Exp Allergy. 1999; 29:633–641.
2. Sadikot RT, Greene T, Meadows KK. Recurrence of primary spontaneous pneumothorax. Thorax. 1997;52:805–809.
25. Schmerber J, Manderlier T. Effect of atmospheric pressure on the occurence of spontaneous pneumothorax. Acta tuberculosea et pneumologica belgica. 1978;69:3–4.
3. Fry WA, Paape K. Pneumothorax. In: Shelds TW, ed. General Thoracic Surgery. Philadelphia: Lea and Febiger; 2000;675–686.
26. Jensen BS. Pneumothorax simplex and atmospheric pressure. Ugeskrift for Laeger. 1980:1603–1606.