Hamilton, Ontario is an industrial city with a population of 300,000 which is situated at the western end of Lake Ontario. Canada's two largest iron and steel mills are located here; the city historically has had relatively poor air quality, which has improved markedly in the last 25 years. Concern about the health effects of current air quality recently led us to carry out an epidemiological study of the effects of air pollution on the respiratory health of over 3500 school children. Respiratory health was measured by pulmonary function testing of each child, and by an assessment of each child's respiratory symptoms via a questionnaire administered to the parents. Previous studies had shown that other environmental factors (e.g. parental smoking, parental cough, socioeconomic level, housing, and gas cooking) might also affect respiratory health, and thus "confound" any potential relationships between health and air pollution. The questionnaire also collected information on many of these confounding factors. For the purposes of initial analysis, the city was divided Into five areas in which differences In air quality were expected. In general, factors which have been associated with poor respiratory health were observed to be more prevalent in areas of poorer air quality.

1. The effect of mental arithmetic tasks on ventilation, breathing pattern, oxygen intake and carbon dioxide output was studied during air breathing and carbon dioxide rebreathing in healthy subjects. 2. Ventilation and breathing frequency increased significantly on performance of the task during 4 min air breathing and 4 min rebreathing; tidal volume was unchanged. The slopes of the ventilatory, frequency and tidal volume responses to carbon dioxide changed little during task performance. 3. During 15 min air breathing, oxygen intake was unchanged with task performance. Carbon dioxide output increased significantly with task performance, as a result of wash-out of carbon dioxide from body stores by the increased ventilation. 4. Mental arithmetic had no effect on the coefficients of variation of the slope and position variables of the ventilatory, frequency and tidal volume responses to carbon dioxide. It is concluded that task performance does not improve the reproducibility of these responses.