The recent article by Nino Kü nzli and colleagues (1), investigat- ing the health effects of the 2003 Southern California wildfires in a group of children, found an ...
Correspondence
629
of granular protein from human eosinophils. Pulm Pharmacol Ther 1997;10:97–104. 2. Ameredes BT, Calhoun WJ. (R)-albuterol for asthma: pro. Am J Respir Crit Care Med 2006;174:965–969.
How Much Smoke Do We Need in Order to Assume That There Is a Fire? To the Editor:
The recent article by Nino Ku¨nzli and colleagues (1), investigating the health effects of the 2003 Southern California wildfires in a group of children, found an association between the environmental levels of particulate matter (PM10) and clinical outcomes, including cough, asthma symptoms, physician visits, and missing school. Surprisingly, there was a stronger association with the duration of the smell of fire. This finding raises a number of concerns. Are the authors implying that future studies should rely on the subjective assessment of the duration of the smell of fire rather than the objective measurement of PM? Furthermore, as the study found that the smell of fire smoke and the level of PM10 were highly correlated, this raises the point of the usefulness of the measurement of PM when simpler and cheaper questionnaires investigating the duration of the smell of smoke are a better predictor of clinical outcomes. The authors acknowledge that the use of questionnaires may be biased in this scenario—could this be the best explanation for the discrepancies above? Even though some of the authors of the current article have previously proved the validity of annoyance scores to assess pollution exposure (2), crucially, this has never been proved for wildfire smoke, which not only is by definition different in its composition from traffic pollution but also occurs acutely and in unusual circumstances. Ku¨nzli and colleagues and the accompanying editorial (3) rightly point out that there may have been a significant spatial variation in smoke concentrations within each community, justifying the fact that the different exposure measures did not coincide. However, I do not see this statement as a confirmation of the validity of the study. On the contrary, this may indicate that neither the measure of smell duration nor PM levels reflected the true exposure to wildfire in this specific study. The fact that this study proves what was expected seems to have been used in the editorial to prove the validity of the conclusions. This is of concern, since a study should be designed to prove or disprove an hypothesis and not the other way around.
From the Authors:
We are grateful for the opportunity to clarify a few issues related to our recent article (1). We do not imply that questionnaire data of fire smoke exposure are better than objective measurements of PM10. However, personal or home outdoor measurements of markers of fire smoke intensity are extremely hard to get and unlikely to be available in fire emergencies. It is for that reason that we had to use questionnaire data. To grade the intensity of exposure, we asked about the number of days fire was smelled in the home. It is biologically plausible that occurrence of symptoms increases not only with the concentration level but with the duration of the exposure to thick fire smoke. While the “dose response” between (reported) duration of fire smell and symptoms is rather convincing, we are well aware—and we extensively discussed—the potential of reporting bias. This concern also originates from our previous research on the association between pollutant levels and reported annoyance. In contrast to the statement in Dr. Paredi’s letter, this association was rather poor on the individual level. It was only on the aggregate group level that annoyance and pollution did correlate, while reported annoyance was determined by a range of individual factors, including sex and health status. Objective measures of PM were only available on the community level. As shown in the online supplement to our article and by Wu and coworkers (Reference 1 of our article), fire smoke concentrations can strongly differ between locations even within the same community. These contrasts in exposure cannot be captured with a single monitor measuring PM10; thus, the objective measurements of pollution come with inherent limitations. Moreover, the mean concentrations across the 5 days with the maximum PM10 would not be expected to correlate with duration of smoke exposure as captured in the questionnaire. Despite the conceptual differences in the two main exposure metrics, namely, reported duration of fire smoke smell and ambient PM10 at the fixed site monitor, associations between symptoms and the two exposure terms were in most, but not all, cases (i.e., not for asthma symptoms or doctor’s visits) rather comparable. While one may thus infer that asthma symptoms and doctor’s visits are more strongly related to the duration of high smoke exposure rather than the absolute level, we believe that this would be an overinterpretation of our data where both exposure metrics had their strength and limitations. While a study can never prove an hypothesis, the null hypothesis of no association between wildfire smoke and a range of symptoms in children is refuted by our results. We have not yet answered the question about major susceptibility factors for these strong effects but will do so in future analyses. Conflict of Interest Statement : None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript.
Conflict of Interest Statement : P.P. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript.
Nino Ku¨nzli Keck School of Medicine University of Southern California Los Angeles, California and ICREA and Institut Municipal de Investigacio Medica Barcelona, Spain Ed Avol John Peters Keck School of Medicine University of Southern California Los Angeles, California
Paolo Paredi National Heart & Lung Institute Imperial College London, United Kingdom References 1. Ku¨nzli N, Avol E, Wu J, Gauderman WJ, Rappaport E, Millstein J, Bennion J, McConnell R, Gilliland F, Berhane K, et al. Health effects of the 2003 Southern California wildfires on children. Am J Respir Crit Care Med 2006;174:1221–1228. 2. Oglesby L, Ku¨nzli N, Monn C, Schindler C, Ackermann-Liebrich U, Leuenberger P. Validity of annoyance scores for estimation of long term air pollution exposure in epidemiologic studies: the Swiss Study on Air Pollution and Lung Diseases in Adults (SAPALDIA). Am J Epidemiol 2000;152:75–83. 3. Vedal S. Where there’s fire, there’s smoke. Am J Respir Crit Care Med 2006;174:1168–1169.
Reference 1. Ku¨nzli N, Avol E, Wu J, Gauderman WJ, Rappaport E, Millstein J, Bennion J, McConnell R, Gilliland F, Berhane K, et al. Health effects of the 2003 Southern California wildfires on children. Am J Respir Crit Care Med 2006;174:1221–1228.
