spirometric data from several thousand randomly selected children in the National Health and Nutrition Survey, which shows that those with greater chronic exposure to ozone experience reductions in lung capacity, as measured by forced vital capacity. An important aspect of most of these studies is that exposure was continuous, but to various levels, rather than dichotomous. This allows one to examine the dose-response relationship and to determine if the pattern suggests a direct linear, monotonic association, or some other dose-response relationship.
In addition to epidemiologic study results (Wagoner et al., 1980) and case reports (Sprince and Kazemi, 1980) among workers exposed to beryllium compounds, neighborhood cases of beryllium disease also have been observed in cross-sectional studies. Chronic beryllium disease is a pulmonary and systemic granulomatous disease caused by inhalation of beryllium. The interval between initial exposure and the clinical manifestation of disease varies from several months to years. Exertional dyspnea is the most common symptom. Other symptoms are cough, fatigue, weight loss, chest pain and arthralgia. Through the study of subjects admitted to the U.S. Beryllium Case Registry, specific criteria for the diagnosis of chronic beryllium disease have been established (Sprince and Kazemi, 1983). In 1948, Hardy reported chronic beryllium disease in persons who lived adjacent to a fluorescent lamp plant in Massachusetts. Additional cases of beryllium disease have been reported among persons living in the vicinity of a beryllium extraction plant (Eisenbud et al., 1949). Atmospheric pollution resulting from stack discharge was thought to be responsible for the latter diseases. By 1960, 47 cases of neighborhood beryllium disease were in the Beryllium Case Registry (Tepper et al., 1961); between 1966 and 1974, 76 new cases were added (Hasan and Kazemi, 1974).
Examples of cross-sectional morbidity studies that involve hazardous-waste sites include one on the Upper Ottawa Street Landfill in Hamilton, Ontario (Hertzman et al., 1987), in which significant differences in symptoms were found between an exposed and a control population, and a study (Baker et al., 1988) of 2039 persons in 606 households located near the Stringfellow Hazardous Waste Disposal Site in California. Both reports contain considerable discussion of how differences in symptom perception and recall can be reduced. A similar study involving a waste site in Lowell, Massachusetts, was reported by Ozonoff et al. (1987). The target population included all households within 400 meters of the site, and the control area was a ring of households between 800 meters and 1200 meters from the site. None of these studies found differences in reproductive out-