A time-series study of air pollution, socioeconomic status, and mortality in Vancouver, Canada

We evaluated the relationship between daily levels of particulate and gaseous phase pollutants and mortality within a dynamic cohort of approximately 550,000 individuals whose vital status was ascertained between 1986 and 1999. Time-series methods were applied to evaluate whether there were differential pollutant effects on daily aggregated numbers of deaths in the cohort that was stratified into quintiles of income as defined by the 1991 and 1996 Canadian censuses. The percent change in all-cause, cardiovascular, respiratory, and cancer daily mortality was calculated in relation to short-term changes in levels of a number of particulate (PM(2.5), PM(10-2.5), total suspended particle co-efficient of haze PM(10), SO(4)) and gaseous (O(3), CO, SO(2), NO(2)) pollutants. The estimated effects of air pollution on mortality were adjusted for day of week effects, and several meteorologic variables including temperature, change in barometric pressure, and relative humidity. Several gaseous pollutants were associated with an increased risk of mortality. Specifically for an increase equivalent to the difference between the 90th and 10th percentiles, the estimated percent change in daily mortality based on the 3-day average of NO(2), and SO(2) was 4.0% and 1.3%, respectively. The corresponding changes in mortality associated with SO(2) were much higher when analyses were restricted to death from respiratory disease. Specifically, a difference between the 90th and 10th percentiles was associated with a 5.6% (95% CI= -0.7% to 12.3%). The daily mean coarse fraction (PM(10-2.5)) was associated with increased cardiovascular mortality (estimated change=5.9%, 95% CI=1.1-10.8%). PM(2.5) was not found to be an important predictor of mortality. For NO(2), CO, and SO(2), there was some suggestion of increased risk of all-cause and cardiovascular mortality at lower levels of socioeconomic status. However, these results should be interpreted cautiously due to the small number of deaths observed within each stratum of socioeconomic status.

Association between particulate- and gas-phase components of urban air pollution and daily mortality in eight Canadian cities

Although some consensus has emerged among the scientific and regulatory communities that the urban ambient atmospheric mix of combustion related pollutants is a determinant of population health, the relative toxicity of the chemical and physical components of this complex mixture remains unclear. Daily mortality rates and concurrent data on size-fractionated particulate mass and gaseous pollutants were obtained in eight of Canada's largest cities from 1986 to 1996 inclusive in order to examine the relative toxicity of the components of the mixture of ambient air pollutants to which Canadians are exposed. Positive and statistically significant associations were observed between daily variations in both gas- and particulate-phase pollution and daily fluctuations in mortality rates. The association between air pollution and mortality could not be explained by temporal variation in either mortality rates or weather factors. Fine particulate mass (less than 2.5 microns in average aerometric diameter) was a stronger predictor of mortality than coarse mass (between 2.5 and 10 microns). Size-fractionated particulate mass explained 28% of the total health effect of the mixture, with the remaining effects accounted for by the gases. Forty-seven elemental concentrations were obtained for the fine and coarse fraction using nondestructive x-ray fluorescence techniques. Sulfate concentrations were obtained by ion chromatography. Sulfate ion, iron, nickel, and zinc from the fine fraction were most strongly associated with mortality. The total effect of these four components was greater than that for fine mass alone, suggesting that the characteristics of the complex chemical mixture in the fine fraction may be a better predictor of mortality than mass alone. However, the variation in the effects of the constituents of the fine fraction between cities was greater than the variation in the mass effect, implying that there are additional toxic components of fine particulate matter not examined in this study whose concentrations and effects vary between locations. One of these components, carbon, represents half the mass of fine particulate matter. We recommend that measurements of elemental and organic carbon be undertaken in Canadian urban environments to examine their potential effects on human health.

Cardiac arrest after administration of Omnipaque radiocontrast medium during endoluminal repair of abdominal aortic aneurysm

We describe a case of acute cardiovascular collapse in a patient with end-stage renal failure undergoing endoluminal repair of an abdominal aortic aneurysm. The 61-yr-old man suffered cardiac arrest shortly after administration of radiocontrast medium (Omnipaque), during deployment of the endovascular device. He had received the same contrast solution for diagnostic angiography on the previous day. He was successfully resuscitated and recovered completely. The differential diagnosis and management are discussed.

Adolescent fears as predictors of depression

BACKGROUND

Recent studies raise questions on the relationship between adolescent fears and risk for major depression.

METHODS

An epidemiologic sample of 776 young people received psychiatric assessments in 1983, 1985-1986, and 1992. Prospective associations were examined between fears in adolescence and future episodes of major depression.

RESULTS

Both overall level of fears and specific fear of dark in adolescence predicted future risk for major depression.

CONCLUSIONS

Relatively high levels of fear in adolescence represent a risk factor for later episodes of major depression.

Associations between daily cause-specific mortality and concentrations of ground-level ozone in Montreal, Quebec

The authors investigated the association between daily variations in ozone and cause-specific mortality. Fixed-site air pollution monitors in Montreal, Quebec, provided daily mean levels of ozone, particles, and other gaseous pollutants. Information on the date and underlying cause of death was obtained for residents of Montreal who died in the city between 1984 and 1993. The authors regressed the logarithm of daily counts of cause-specific mortality on mean levels of ozone, after accounting for seasonal and subseasonal fluctuations in the mortality time series, non-Poisson dispersion, and weather variables. The effect of ozone on mortality was generally higher in the warm season and among persons aged 65 years or over. For an increase in the 3-day running mean concentration of ozone of 21.3 microg/m(3), the percentage of increase in daily deaths in the warm season was the following: nonaccidental deaths, 3.3% (95% confidence interval (CI): 1.7, 5.0); cancer, 3.9% (95% CI: 1.0, 6.91); cardiovascular diseases, 2.5% (95% CI: 0.2, 5.0); and respiratory diseases, 6.6% (95% CI: 1.8, 11.8). These results were independent of the effects of other pollutants and were consistent with a log-linear response function.

Identification of persons with cardiorespiratory conditions who are at risk of dying from the acute effects of ambient air particles

This study was undertaken to identify subgroups of the population susceptible to the effects of ambient air particles. Fixed-site air pollution monitors in Montreal, Quebec, Canada, provided daily mean levels of various measures of particulates and gaseous pollutants. Total sulfates were also measured daily (1986-1993) at a monitoring station 150 km southeast of the city (Sutton, Quebec, Canada). We used coefficient of haze (COH), extinction coefficient, and Sutton sulfates to predict fine particles and sulfates from a fine particles model for days that were missing. We used the universal Quebec medicare system to obtain billings and prescriptions for each Montreal resident who died in the city from 1984 to 1993. These data were then used to define cardiovascular and respiratory conditions that subjects had before death. Using standard Poisson regression time-series analyses, we estimated the association between daily nonaccidental mortality and daily concentrations of particles in the ambient air among persons with cardiovascular and respiratory conditions diagnosed before death. We found no persuasive evidence that daily mortality increased when ambient air particles were elevated for subgroups of persons with chronic upper respiratory diseases, airways disease, cerebrovascular diseases, acute coronary artery disease, and hypertension. However, we found that daily mortality increased linearly as concentrations of particles increased for persons who had acute lower respiratory diseases, chronic coronary artery diseases (especially in the elderly), and congestive heart failure. For this latter set of conditions, the mean percent increase in daily mortality (MPC) for an increase in the COH across its interquartile range (18.5 COH units per 327.8 linear meters), averaged over the day of death and the 2 preceding days, was MPC = 5.09% [95% confidence interval (CI) 2.47-7.79%], MPC = 2.62 (95% CI 0.53-4.75%), and MPC = 4.99 (95% CI 2.44-7.60%), respectively. Adjustments for gaseous pollutants generally attenuated these associations, although the general pattern of increased daily mortality remained. In addition, there appeared to be a stronger association in the summer season. The positive associations found for persons who had acute lower respiratory diseases and congestive heart failure are consistent with some prevailing hypotheses and may also be consistent with recent toxicologic data implicating endothelins. Further epidemiologic studies are required to confirm these findings.

The association between daily mortality and ambient air particle pollution in Montreal, Quebec. 1. Nonaccidental mortality

This study was undertaken to determine whether variations in concentrations of particles in the ambient air of Montreal, Quebec, during the period 1984 to 1993, were associated with daily variations in nonaccidental mortality. Fixed-site air pollution monitors in Montreal provided daily mean levels of various measures of particulates and gaseous pollutants. Total sulfates were also measured daily (1986-1993) at a monitoring station 150 km southeast of the city (Sutton, Quebec). We estimated associations for PM(2.5), PM(10), total suspended particles, coefficient of haze (COH), extinction coefficient, and sulfates. We used coefficient of haze, extinction coefficient, and Sutton sulfates to predict fine particles and sulfates for days that were missing. To estimate the associations between nonaccidental mortality and ambient air particles, we regressed the logarithm of daily counts of nonaccidental mortality on the daily mean levels for the above measures of particulates, after accounting for seasonal and subseasonal fluctuations in the mortality time series, non-Poisson dispersion, weather variables, and gaseous pollutants. There were 140,939 residents of Montreal who died during the study period. We found evidence of associations between daily nonaccidental deaths and most measures of particulate air pollution. For example, the mean percentage increase (MPC) for an increase of total suspended particles of 28.57 microg/m(3) (interquartile range, IQ), evaluated at lag 0 days, was 1.86% (95% confidence interval (CI): 0.00-3.76%), and for an increase of coefficient of haze (IQ=18.5 COH units per 327.8 linear m) the MPC was 1.44% (95% CI: 0.75-2.14%). These results are similar to findings from other studies (the mean percentage increase in nonaccidental deaths for a 100 microg/m(3) increase in daily total suspended particles was 6.7%). We also found increases for fine particles and for inhalable particles, but the confidence intervals included unity. All measures of sulfates showed increased daily mortality; e.g., the MPC for sulfates from fine particles (IQ=3.51 microg/m(3)) was 1.86% (95% CI: 0.40-3.35%). We generally found higher excesses in daily mortality for persons 65 years of age and for exposures averaged across lags 0, 1, and 2 days. The slope of the association between daily mortality and ambient air particles in Montreal, which has lower levels of pollution than most major urban centers, is similar to that reported in most other industrialized cities. This study therefore provides further evidence that the association is linear and that any threshold effect, should it exist, would be found at lower levels of air pollution than those found in Montreal.

The association between daily mortality and ambient air particle pollution in Montreal, Quebec. 2. Cause-specific mortality

This study was undertaken to determine whether variations in concentrations of particulates in the ambient air of Montreal, Quebec, during the period 1984 to 1993, were associated with daily variations in cause-specific daily mortality. Fixed-site air pollution monitors in Montreal provided daily mean levels of various measures of particles and gaseous pollutants. Total sulfate was also measured daily (1986-1993) at a monitoring station 150 km southeast of the city (Sutton, Quebec). We used coefficient of haze (COH), extinction coefficient, and sulfate from the Sutton station to predict fine particles and sulfate from fine particles for days that were missing. We estimated associations between cause-specific mortality and PM(2.5), PM(10), predicted fine particles and fine sulfate particles, total suspended particles, coefficient of haze, extinction coefficient, and total sulfate measured at the Sutton station. We selected a set of underlying causes of death, as recorded on the death certificates, as the endpoint and then regressed the logarithm of daily counts of cause-specific mortality on the daily mean levels for the above measures of particulates, after accounting for seasonal and subseasonal fluctuations in the mortality time series, non-Poisson dispersion, weather variables, and gaseous pollutants. We found positive and statistically significant associations between the daily measures of ambient particle mass and sulfate mass and the deaths from respiratory diseases and diabetes. The mean percentage change in daily mortality (MPC), evaluated at the interquartile range for pollutants averaged over the day of death and the preceding 2 days, for deaths from respiratory diseases was MPC(COH)=6.90% (95% CI: 3.69-10.21%), MPC(Predicted PM2.5)= 9.03% (95% CI: 5.83- 12.33%), and MPC(Sutton sulfate)=4.64% (95% CI: 2.46-6.86%). For diabetes, the corresponding estimates were MPC(COH)=7.50% (95% CI: 1.96-13.34%), MPC(Predicted PM2.5)=7.59% (95% CI: 2.36-13.09%), and MPC(Sutton sulfate)=4.48% (95% CI: 1.08-7.99%). Among individuals older than 65 years at time of death, we found consistent associations across our metrics of particles for neoplasms and coronary artery diseases. Associations with sulfate mass were also found among elderly persons who died of cardiovascular diseases and of lung cancer. These associations were consistent with linear relationships. The associations found for respiratory diseases and for cardiovascular diseases, especially in the elderly, are in line with some of the current hypotheses regarding mechanisms by which ambient particles may increase daily mortality. The positive associations found for cancer and for diabetes may be understood through a general hypothesis proposed by Frank and Tankersley, who suggested that persons in failing health may be at higher risk for external insults through the failure of regulating physiological set points. The association with diabetes may be interpreted in light of recent toxicological findings that inhalation of urban particles in animals increases blood pressure and plasmatic levels of endothelins that enhance vasoconstriction and alter electrophysiology. Further research to confirm these findings and to determine whether they are causal is warranted.

Adolescent depressive symptoms as predictors of adult depression: moodiness or mood disorder?

OBJECTIVE

The authors' goal was to examine the relationship between subclinical depressive symptoms in adolescence and major depressive episodes in adulthood.

METHOD

An epidemiologic sample of 776 young people received psychiatric assessments in 1983, 1985, and 1992. Among adolescents not meeting criteria for major depression, the authors estimated the magnitude of the association between subclinical adolescent depressive symptoms and adult major depression.

RESULTS

Symptoms of major depression in adolescence strongly predicted an adult episode of major depression: having depressive symptoms more than two-standard-deviations above the mean in number predicted a two-fold to three-fold greater risk for an adult major depressive episode.

CONCLUSIONS

Symptoms of depression in adolescence strongly predict an episode of major depression in adulthood, even among adolescents without major depression.

Prospective associations between somatic illness and mental illness from childhood to adulthood

The association between somatic illness and psychiatric illness is well established in adults but is less clear in childhood and adolescence. A cohort of over 700 randomly selected children in Upstate New York were studied from ages 1-10 years in 1975 to young adulthood in 1992. Psychiatric and physical health were assessed by means of follow-up youth and parent interviews at 8-, 2 1/2-, and 6-year intervals (in 1983, 1985-1986, and 1991-1993). Cross-sectional and longitudinal analyses investigated: 1) the consistency of the relation between physical illness and mental illness in childhood; 2) the specificity of major depressive disorder (MDD) in accounting for the relation; 3) the specificity of immunologically mediated medical disorders in this relation; and 4) whether this relation was attributable to risks associated with low socioeconomic status. Cross-sectionally, ill health was associated with increased risk of psychiatric disorders at all ages, with significant odds ratios (ORs) ranging from 1.76 to 3.26. In prospective analyses, ill health increased the risk of new-onset MDD at all ages (ORs = 2.05-4.48). MDD also predicted subsequent ill health, independent of prior health problems (ORs = 3.81 and 4.04). Relations were not attributable to familial socioeconomic status. Associations were particularly strong between MDD and medical disorders associated with alterations in immunologic factors (ORs = 1.83-6.41). Theories of common immune-mediated vulnerabilities to medical illness and depression are consistent with these associations.

The risk for early-adulthood anxiety and depressive disorders in adolescents with anxiety and depressive disorders

BACKGROUND

Various studies find relationships among anxiety and depressive disorders of adolescence and adulthood. This study prospectively examines the magnitude of longitudinal associations between adolescent and adult anxiety or depressive disorders.

METHODS

An epidemiologically selected sample of 776 young people living in upstate New York received DSM-based psychiatric assessments in 1983, 1985, and 1992 using structured interviews. The magnitude of the association between adolescent and adult anxiety or depressive disorders was quantified using odds ratios generated from logistic regression analyses and from a set of latent Markov analyses. We focus on longitudinal associations among narrowly defined DSM anxiety or depressive disorders.

RESULTS

In simple logistic models, adolescent anxiety or depressive disorders predicted an approximate 2- to 3-fold increased risk for adulthood anxiety or depressive disorders. There was evidence of specificity in the course of simple and social phobia but less specificity in the course of other disorders. Results from the analyses using latent variables suggested that while most adolescent disorders were no longer present in young adulthood, most adult disorders were preceded by adolescent disorders.

CONCLUSIONS

An anxiety or depressive disorder during adolescence confers a strong risk for recurrent anxiety or depressive disorders during early adulthood. Most anxiety and depressive disorders in young adults may be preceded by anxiety or depression in adolescence.

Psychiatric symptoms in adolescence as predictors of obesity in early adulthood: a longitudinal study

OBJECTIVES

This study examined the longitudinal relationship between psychopathology and obesity in young adulthood.

METHODS

More than 700 youth in a population-based sample were psychiatrically assessed in 1983 (mean age = 14 years) and 1992 (mean age = 22 years). Self-reported body mass index (BMI) in 1992 was regressed on measures of depression and conduct disorder as well as a set of covariates including indices of physical health, social class, intelligence, and cigarette and alcohol use. Associations were examined with BMI treated as a continuous variable and with a binary index of obesity derived from the BMI distribution in each gender.

RESULTS

BMI in young adults was positively related to a number of covariates. With all covariates controlled, BMI was inversely related to adult depressive symptoms in males but not females. BMI was positively related to adolescent symptoms of conduct disorder in both sexes. Similar associations were found between psychiatric symptoms and obesity.

CONCLUSIONS

Conduct disorder symptoms in adolescence predicted BMI and obesity in early adulthood. These associations remained after controlling for factors that can affect the association between psychopathology and obesity.

The association between major depression and headache: results of a longitudinal epidemiologic study in youth

Retrospective epidemiologic research in adults suggests that a long-term association between major depression and headache arises during childhood or adolescence. This study uses data from a prospective epidemiologic study to examine the association between major depression and headache from late childhood into early adulthood. An epidemiologically selected sample of 776 youth, aged 9-18, was assessed psychiatrically in 1983 using DISC interviews with both youth and parent informants. Reassessments were conducted in 1985-1986 and 1992. Current and past histories of functionally impairing migraine or chronic headache were elicited in both 1985-1986 and 1992. Regression analyses examined the relationships between major depression and headache status. The prevalence of current functionally impairing headache was approximately 10% in both 1985-1986 and 1992. There were lifetime and cross-sectional associations between headache and major depression. Headache was approximately twice as common in depressed adolescents compared with nondepressed adolescents. Major depression in adolescents, without current or past headache, prospectively predicted the new onset of headaches in young adulthood. Among adolescents who had no history of chronic impairing headache in 1985-1986, those with current major depression faced a nearly tenfold increased risk of developing such headaches at some time during the next 7 years. Consistent with findings of retrospective studies among adults, a longitudinal/ developmental relationship between major depression and functionally impairing headache was found in this prospective epidemiologic study of youth. These findings suggest that (1) neurochemical and pharmacologic commonalities between depression and headache should receive further investigation and that (2) it may be clinically useful to inquire about family history of headache syndromes in adolescents with major depression because such questioning may provide insight about the risk of subsequent functionally impairing headache in a depressed adolescent.

Refugees' and immigrants' mental health: association of demographic and post-immigration factors

Several demographic and post-immigration factors related to self-reported symptoms of anxiety and depression of immigrants in New Zealand were examined in a study involving a sample of 129 Southeast Asian refugees, 57 Pacific Island immigrants, and 63 British immigrants to New Zealand. A questionnaire and the Hopkins Symptom Checklist-25 (HSCL-25) in English and in three Southeast Asian translations were administered, followed by an interview in English or with an interpreter. The findings suggest that demographic characteristics are not associated with symptom level. Post-immigration factors, such as experiencing discrimination in New Zealand, not having close friends, being unemployed, and spending most of one's time with one's own ethnic group affected anxiety and depression scores.

Discriminating depression and anxiety in youth: a role for diagnostic criteria

OBJECTIVE

To test the hypothesis that anxiety and depression in youth, as in adults, become increasingly discriminable when youth meet criteria for an emotional disorder.

METHOD

The study uses cross-sectional data at two points in time from a large (n = 776) community sample of youths, aged eight to twenty. Associations between major depression and five anxiety disorders (overanxious, obsessive compulsive and separation anxiety disorders, and social and simple phobias) are examined by symptom scale and diagnosis.

RESULTS

Anxiety and depression are moderately correlated, and substantially comorbid by diagnostic category. Symptoms are more discriminable among youths with diagnoses of at least one emotional disorder than among those without. A single factor accounts for symptoms among the non-diagnosed but multiple factors are required for the diagnosed group.

CONCLUSIONS

Anxiety and depression are discriminable among youth who meet criteria for a specific emotional disorder but more highly associated among youths without such a diagnosis. This suggests that in youth, as has been shown in adults, depression and anxiety become increasingly discriminable as emotional psychopathology becomes more severe.

Emotional problems during youth as predictors of stature during early adulthood: results from a prospective epidemiologic study

OBJECTIVE

Adults with emotional disorders exhibit abnormalities in growth hormone secretion. If these abnormalities were to occur during childhood, they could affect growth. The purpose of this study was to examine the relationship between youth emotional disorders and stature in early adulthood.

METHODS

Using data from a prospective epidemiologic study of youth psychopathologic status, we used linear regression to examine the prospective relationship between anxiety disorders (separation anxiety and over-anxious disorders) or major depressive disorder in youth and stature in early adulthood.

RESULTS

Anxiety disorders during childhood prospectively predicted relatively short stature in early adulthood among females, accounting for more than 5% of the variance in adult height. However, these associations were not found among males.

CONCLUSIONS

There may be an association between abnormalities in growth and emotional problems in youth. Further research should examine biological measures related to growth among youth with emotional disorders.

The mental health pattern of migrants: is there a euphoric period followed by a mental health crisis?

Sluzki's 1986 mental health model of the migratory process was tested with migrants (both refugees and immigrants), to New Zealand. Its central feature, suggesting an initial symptom free and euphoric phase after arrival in the country of settlement, followed by a crisis stage, was examined for 129 Southeast Asian refugees, 57 Pacific Island immigrants and 63 British immigrants. A questionnaire and the Hopkins Symptom Checklist-25, in English and in three Southeast Asian language translations, were administered face-to-face. All respondents had arrived in New Zealand within the last 15 years. The findings did not support Sluzki's model. Refugees and immigrants in the group with less than six months of residence were not symptom free. Neither did the group with six months to six years residence demonstrate a deterioration in mental health. However, mean depression levels were slightly lower for those who had lived in New Zealand for over six years, suggesting that mental health may improve the longer both refugees and immigrants reside in the host country.

Hemolytic uremic syndrome in New Jersey

In an 18-month study period, the authors identified 23 cases of hemolytic uremic syndrome (HUS) in New Jersey children. Increased identification and reporting of Escherichia coli O157:H7 are needed to prevent additional morbidity and mortality.

Prevalence of childhood asthma across Canada

BACKGROUND

A large cross-sectional study provided an opportunity to estimate the prevalence of childhood asthma in several regions across Canada.

METHODS

In 1988, approximately 18,000 questionnaires were distributed to the families of 5-8 year old children in 30 communities from the following six regions across Canada: the interior of British Columbia, southeastern Saskatchewan, southwestern Ontario, the central region of Ontario, southern Quebec, and the Maritimes (Nova Scotia and Prince Edward Island). These communities were free of point-source air pollutants and selected to represent a range of ambient sulphate concentrations. In all 14,948 questionnaires were returned representing an 83% response rate.

RESULTS

Currently present, physician-diagnosed asthma was reported for 4.7% of children by their parents. Persistent wheezing was reported for 13% and persistent cough for 5.9%. Asthma was most common in the two Maritime provinces (7.4%), and least common in British Columbia (3.3%) and Quebec (3.4%). Similar regional differences were seen for persistent cough, persistent wheeze, and also hospital separation rates for asthma which were approximately 800 per 100,000 for the Maritimes and 396 per 100,000 for British Columbia. Differences persisted despite adjustments for several host and environmental (indoor and outdoor) characteristics.

CONCLUSIONS

The east coast of Canada may be an endemic area of asthma in Canada. If confirmed by objective measures of asthma, a detailed aetiologic investigation could enhance understanding of this phenomenon and the major environmental determinants of asthma morbidity in general.

Effects of low ambient levels of ozone and sulfates on the frequency of respiratory admissions to Ontario hospitals

To investigate the acute respiratory health effects of ambient air pollution, the number of emergency of urgent daily respiratory admissions to 168 acute care hospitals in Ontario were related to estimates of exposure to ozone and sulfates in the vicinity of each hospital. Ozone levels were obtained from 22 monitoring stations maintained by the Ontario Ministry of the Environment for the period January 1, 1983 to December 31, 1988. Daily levels of sulfates were recorded at nine monitoring stations representing three different networks operated by the Ontario Ministry of the Environment and Environment Canada. Positive and statistically significant associations were found between hospital admissions and both ozone and sulfates recorded on the day of admission and up to 3 days prior to the date of admission. Five percent of daily respiratory admissions in the months of May to August were associated with ozone, with sulfates accounting for an additional 1% of these admissions. Ozone was a stronger predictor of admissions than sulfates. Positive and statistically significant (P < 0.05) associations were observed between the ozone-sulfate pollution mix and admissions for asthma, chronic obstructive pulmonary disease, and infections. Positive associations were also found in all age groups, with the largest impact on infants (15% of admissions associated with the ozone-sulfate pollution mix) and the least effects on the elderly (4%). Temperature had no effect on the air pollution-admission relationship. Ozone (lagged 1 day) and sulfates (lagged 1 day) displayed a positive association with respiratory admissions for 91 and 100% of the 168 acute care hospitals, respectively. Air pollution was not related to a class of nonrespiratory admissions, which served as a negative control, nor was it related to admissions in the winter months of December to March, when ozone and sulfate levels are low and when people spend a considerable amount of time indoors.