An investigation of inhaled ozone dose and the magnitude of airway inflammation in healthy adults.

Exposure to ozone has been associated in epidemiologic and field studies with a range of health outcomes: increased daily mortality (1); impaired lung function (2–4); and inflammation and epithelial injury in recreational runners (5) and cyclists (6), as well as increased hospital admissions for preexisting respiratory disease (7, 8). A biological basis for these responses has been provided by findings from animal (9) and human laboratorybased studies (10), where transient decrements in lung function, increased hyperreactivity, and airway inflammation have all been demonstrated after exposure to environmentally relevant concentrations of ozone (11, 12). Although the database for ozoneinduced health effects is considerable, a number of issues remain unresolved. The relationship between inhaled ozone dose and the magnitude of the resultant pulmonary response has been limited to a consideration of transient lung function decrements (12–14), as these measurements are noninvasive and easy to perform. Similar analyses addressing the relationship between ozone dose and acute pulmonary inflammation and tissue injury have not been performed. Here, the available database is more limited, and issues related to variation in study design, airway sampling, and the endpoints studied have made a similar approach more problematic. Establishing these relationships is vital in determining threshold doses of ozone below which adverse responses are negligible in the healthy population. As the magnitude of lung function decrements is not simplistically related to epithelial injury or inflammation (15), one cannot rely on estimates derived from these modeling studies to determine threshold values for these other biologically important endpoints. Here, we describe the first attempt to model the ozone dose– inflammatory response based on a metaanalysis of data from published human chamber studies, where airway responses were assessed using bronchoscopy-based lavage. Analyses of inflammatory responses were based on the percentage of neutrophils recovered by bronchoscopy-based lavage in response to a set ozone dose, expressed as a function of concentration (mg/m), Ve (L/minute/m body surface area), and exposure time (minutes) CVT (mg/m body surface area). Analysis of altered permeability responses was more complex and relied on the fold increase in