Australian Child Health and Air Pollution Study

AUSTRALIAN CHILD HEALTH AND AIR POLLUTION STUDY

(ACHAPS)

FINAL REPORT

May 2012

Format of the report

Executive summary

Part A: Study background and methods

Part B: The cross-sectional study : chronic exposure to air pollution: effects upon children’s health

Part C: The panel study: acute exposure to air pollution: effects upon children’s health

Part D: Conclusions

Part E: References

Appendices

Acknowledgements

We are very grateful to the school principals who generously agreed to their school taking part in this study, the parents who gave their time and consent for their children to be involved, and the children who participated, especially those who were part of the panel study.

We would also like to thank the Australian Research Council (ARC linkage grant # LP0562551) for their financial support, and local environmental protection authorities for their in-kind contributions which made the execution of this study possible.

Specifically, we would like to acknowledge contributions from:

·  David Powell and Des Clayton from the Environment Protection Authority in the ACT

·  Anne Marie McCarthy, Susan Lloyd, Alistair Nairn, Sue Gipson, Susana Young, Gary Laidlaw and Margot Finn from the Environment Protection Authority in Victoria

·  Emma Clarke, Phil Kingston, Scott McDowall and Susie Kolb from the Environmental Protection Agency in Queensland

·  Jason Caire and Polly Weckert from the Environment Protection Authority in South Australia

·  Karl Carrabotta and Jodie Bell from the Department of Environment and Conservation in Western Australia

·  Trevor Solomon, Nigel Routh, Cheryl Palmer, Debbie Maddison, Pamiela Berenson, Michael Johnson and Hiep Duc from the Department of Environment, Climate Change and Water in New South Wales.

In addition, we would like express our gratitude to members of the epidemiology department at the Woolcock Institute of Medical Research who gave their time and experience in order to support the fieldwork data collection: Wafaa Nabil Ezz, Elena Belousova, Kity Ng, Brett Toelle, and Tessa Bird. We would also like to acknowledge the successful completion of the panel study follow-up by Adrian Forero, Paula Garay and Lucy Marks.

Finally, a special thank you to: Amanda Dawes for her support with the graphic design of study forms and for setting up the study website; to Lucy Williams for her support with the study media releases; to Aaron Kelsey for his technical support with the nitric oxide equipment; to Elena Belousova and Robert Li who collaborated with the database creation and management; and to the University of Queensland staff who were involved in entering the data – Tania Patrao, Shannon Dias, Rumna De, Helder-Fernando Ntimane, Lu Jie, and Aishath Niyaf.

Project Management Committee

Principal Investigator / Professor Gail Williams
University of Queensland
Co-investigators / Professor Guy B Marks, Head of Epidemiology, Woolcock Institute of Medical Research
Dr Lyn Denison
Environment Protection and Heritage Council / Air Quality EPA Victoria
Professor Bin Jalaludin
Centre for Research, Evidence Management and Surveillance, Sydney South West Area Health Service
EPHC Project Manager / Kerry Scott
Environment Protection and Heritage Council / NEPC Service Corporation
Study Coordinator / Adriana M Cortés
Woolcock Institute of Medical Research
Data analysis and
report preparation / Professor Gail Williams
Professor Bin Jalaludin
Adriana M Cortés
Data management / Mr Robert Li, University of Queensland
Dr Adrian Barnett, Queensland University of Technology
Data collection / Woolcock Institute of Medical Research
Adriana M Cortés
Kate M Hardaker
Brett G Toelle
Paola T Espinel

Shortened forms

ACHAPS / Australian Child Health and Air Pollution Study
AQM / air quality monitor
BMI / body mass index
Bsp / the measure of back light scattering particles or suspended fine particles measured by nephelometry
CAPS / Childhood Asthma Prevention Study
CI / confidence interval
CO / carbon monoxide
CO2 / carbon dioxide
eNO / nitric oxide in the exhaled breath
EPA / Environmental Protection Agency
EPHC / Environment Protection and Heritage Council
FEF25-75% / average of expired flow over the middle half of forced vital capacity
FEV1 (litres) / forced expiratory volume in one second
FEV1/VC / FEV1 as a percentage of vital capacity or forced vital capacity
FVC (litres) / forced vital capacity
HR / hazard ratio
IQR / interquartile range
ISSAC / International Study of Asthma and Allergies in Childhood
µg/m3 / micrograms per cubic meter
MMEF / maximal midexpiratory flow
NEPC / National Environment Protection Council
(the) NEPM / National Environment Protection (Ambient Air Quality) Measure
NO / nitric oxide
NO2 / nitrogen dioxide
NOX / oxides of nitrogen (NO + NO2)
OR / odds ratio
O3 / ozone
PAH / polycyclic aromatic hydrocarbon
PEF / peak expiratory flow
PEFR / peak expiratory flow rate
PM / particulate matter
PM10 / particulate matter less than 10 μm in diameter
PM2.5 / particulate matter less than 2.5 μm in diameter
ppb / parts per billion
ppm / parts per million
RSP / respirable suspended particulate
SCCHS / Southern California Children’s Health Study
SO2 / sulphur dioxide
SD / standard deviation
SEIFA / socio-economic indexes for areas
SPFR / standardised peak flow rates
TSP / total suspended particulate matter

TABLE OF CONTENTS

Executive summary i

Part A: STUDY BACKGROUND AND METHODS 1

1. INTRODUCTION 2

1.1. Purpose 2

1.2. Background 2

1.2.1. Australian air quality standards 2

1.2.2. Relevance of standards to Australia 4

1.2.3. Mechanisms affecting human health 4

1.2.4. Exposure measurement 4

1.2.5. Children are not ‘little adults’ 5

1.2.6. Effects of air pollution on child health 5

1.2.7. Australian evidence 6

2. STUDY AIMS AND APPROACH 8

2.1. Background 8

2.2. Overall aims of this study 10

2.3. Approach 10

3. REVIEWS AND CLEARANCES 11

3.1. University ethics approvals 11

3.2. Departments of education approval 11

3.3. Amendments to ethics approval 12

3.4. Application to the Catholic Education Office 12

3.5. Working with Children Check 12

4. OVERALL STUDY DESIGN 13

4.1. Site selection 13

4.2. Air pollution exposures 13

4.3. Sample size and power 13

5. LOGISTICS 14

5.1. Fieldwork planning 14

5.2. EPA staff support 14

5.3. Time frame for cross-sectional and panel study 15

Part B: CROSS-SECTIONAL STUDY 17

1. STUDY DESIGN AND MEASUREMENT 18

1.1. Aims 18

1.2. Data collection 18

1.2.1. Questionnaire 18

1.2.2. Clinical assessments 19

1.3. Statistical analysis 21

2. RECRUITMENT 23

2.1. School recruitment 23

2.2. Participant recruitment 23

3. RESULTS 25

3.1. Demographic characteristics of cross-sectional study participants 25

3.2. Respiratory conditions of cross-sectional study participants 26

3.3. Lung function 29

3.4. Household environment 29

3.5. Air pollutant exposures 31

3.6. SEIFA characteristics 35

3.7. Pollutants and lung function 36

3.7.1. Single pollutant models 36

3.7.2. Joint pollutant models 41

3.7.3. Interaction models 48

3.8. Pollutants and respiratory symptoms 51

3.8.1. Single pollutant models 51

3.8.2. Joint pollutant models 56

3.8.3. Interaction models 58

4. DISCUSSION 60

4.1. Summary of findings 60

4.2. Comparison with other studies of chronic exposure in school children 60

4.3. Methodological issues 69

Part C: PANEL STUDY 71

1. STUDY DESIGN AND MEASUREMENT 72

1.1. Aims and objectives 72

1.2. Methods 72

1.2.1. Participant selection 72

1.2.2. Panel study measurements 72

1.2.3. Panel study adherence maintenance procedures 74

1.3. Statistical analyses 75

2. RESULTS 76

2.1. Demographic and baseline data 76

2.2. Diary data 77

2.3. Air pollution data 83

2.4. Air pollutants and lung function, symptoms and medication use 85

2.4.1. Single pollutant models 85

2.4.2. Ozone: warm season 98

2.4.3. SO2: excluding children from Port Pirie 102

2.4.4. SO2: two pollutant models with PM10 106

2.4.5. NO2: two pollutant models with ozone 110

2.4.6. NO2: in children with unflued gas heating in the home 122

2.4.7. PM10: two pollutant models 126

3. DISCUSSION 143

3.1. Single pollutant models 143

3.2. Two pollutant models 144

3.3. Restricted models 144

3.4. Comparisons with other studies 145

3.4.1. Particulate matter 145

3.4.2. Ozone 147

3.4.3. Nitrogen dioxide 148

3.4.4. Sulphur dioxide 150

3.4.5. Carbon monoxide 151

3.5. Methodological issues 152

3.5.1. Measurement of lung function 152

3.5.2. Fixed site air pollution monitoring 152

3.5.3. Accuracy of diary keeping 153

3.5.4. Learning effect 153

3.5.5. Hours spent outdoors doing vigorous physical activity 153

3.5.6. Effects of temperature 154

3.5.7. Single-pollutant versus multi-pollutant models 154

3.5.8. Collinearity among the independent variables 154

3.5.9. Multiple comparisons 155

Part D: CONCLUSIONS 156

Part E: REFERENCES 160

LIST OF TABLES

A: STUDY BACKGROUND AND METHODS

Table 1.1. Australian air quality standards. 3

Table 3.1. Departments of education submission and approval dates. 11

Table 5.1. Fieldwork dates for 2007. 15

Table 5.2. Fieldwork dates for 2008. 15

B: CROSS-SECTIONAL STUDY

Table 2.1. Description of test completion in ACHAPS. 24

Table 2.2. Reasons for tests not completed in ACHAPS. 24

Table 3.1.Child characteristics (N = 2,860); numbers and per cent unless otherwise specified. 25

Table 3.2. History of asthma; numbers (n) and per cent (%). 26

Table 3.3. History of wheezing; numbers (n) and per cent (%). 26

Table 3.4. History of cough in the last 12 months; numbers (n) and per cent (%). 26

Table 3.5. History of itchy rash/eyes or rhinitis; numbers (n) and per cent (%). 27

Table 3.6. History of diagnosed illnesses ever; numbers (n) and per cent (%) 27

Table 3.7. Child allergies; numbers (n) and per cent (%). 27

Table 3.8. Family history of asthma or allergies; numbers (n) and per cent (%). 28

Table 3.9. Child anthropometry; mean, standard deviation (SD), and range. 28

Table 3.10. Child lung function tests; mean, standard deviation (SD), and range. 29

Table 3.11. Household exposures; numbers (n) and per cent (%). 30

Table 3.12. Smoking exposure of child; numbers (n) and per cent (%). 31

Table 3.13. Other exposures: numbers (n) and per cent (%). 31

Table 3.14. Numbers of sites available for pollutant measures. 31

Table 3.15. Mean and variability of lifetime child exposures. 32

Table 3.16. Mean and variability of recent child exposures. 32

Table 3.17. Pearson correlations among lifetime child exposures . 32

Table 3.18. Summary of correlations among child exposures. 33

Table 3.19. Mean and variability of SEIFA indices of child's postcode of residence. 35

Table 3.20. Correlations among child exposures and SEIFA indices. 36

Table 3.21. Lung function: single pollutant models. Lifetime exposure. 38

Table 3.22. Lung function: single pollutant models. Recent exposure. 39

Table 3.23. Lifetime exposure: effects per unit pollutant, all schools. 42

Table 3.24. Recent exposure: effects per unit pollutant, all schools. 42

Table 3.25. Lifetime exposure: effects per unit pollutant, 10 high-O3 schools omitted. 43

Table 3.26. Recent exposure: effects per unit pollutant, 10 high-O3 schools omitted. 43

Table 3.27. Lifetime exposure: effects per unit pollutant. 45

Table 3.28. Recent exposure: effects per unit pollutant. . 46

Table 3.29. Lifetime exposure: effects per unit pollutant 47

Table 3.30. Recent exposure: effects per unit pollutant. 47

Table 3.31. Modification of SO2 (ppb) effects by atopic status. 49

Table 3.32. Modification of SO2 (ppb) effects by gender. 50

Table 3.33. Symptoms: single pollutant models, lifetime exposurel. 53

Table 3.34. Symptoms: single pollutant models, recent exposure;. 54

Table 3.35. Lifetime exposure: effects per unit pollutant; unrestricted O3. 56

Table 3.36. Recent exposure: effects per unit pollutant: unrestricted O3. 57

Table 3.37. Lifetime exposure: effects per unit pollutant: restricted O3. 57

Table 3.38. Recent exposure: effects per unit pollutant: restricted O3 58

Table 3.39. Modification of PM2.5 (mg/m3) effects by atopic status 59

PANEL STUDY

Table 2.1. Demographic and clinical characteristics of subjects 77

Table 2.2. Timing of panel study in each jurisdiction 78

Table 2.3. Available diary days per child for selected outcome variables 79

Table 2.4. Summary statistics for selected continuous outcome variables 80

Table 2.5. Number of children by percentage of diary days with night symptoms 81

Table 2.6. Number of children by percentage of diary days with day time symptoms 82

Table 2.7. Number of children by percentage of diary days with medication use 83

Table 2.8. Associations between air pollutants and lung function 86

Table 2.9. Associations between air pollutants and night symptoms and medication use 88

Table 2.10. Associations between air pollutants and day time symptoms 92

Table 2.11. Associations between air pollutants and day time medication use 96

Table 2.12. Associations between warm season ozone and lung function 98

Table 2.13. Associations between warm season ozone and night symptoms and medication 99

Table 2.14. Associations between warm season ozone and day time symptoms 100

Table 2.15. Associations between warm season ozone and day time medication use 101

Table 2.16. Associations between SO2 and lung function, excluding Port Pirie children 102

Table 2.17. Associations between SO2 and night symptoms, excluding Port Pirie children 103

Table 2.18. Associations between SO2 and day symptoms, excluding Port Pirie children 104

Table 2.19. Associations between SO2 and day medication use, excluding Port Pirie children 105

Table 2.20. Associations between SO2 and lung function in two pollutant models with PM10 106

Table 2.21. SO2 and night symptoms and medication use in two pollutant models with PM10 107

Table 2.22. SO2 and day time symptoms in two pollutant models with PM10 108

Table 2.23. SO2 and day time medication use in two pollutant models with PM10 109

Table 2.24. NO2 and lung function in two pollutant models with 1-hour ozone 110

Table 2.25. NO2 and night symptoms in two pollutant models with 1-hour ozone 111

Table 2.26. NO2 and day time symptoms in two pollutant models with 1-hour ozone 112

Table 2.27. NO2 and day time medication use in two pollutant models with 1-hour ozone 113

Table 2.28. NO2 and lung function in two pollutant models with 4-hour ozone 114

Table 2.29. NO2 and night symptoms in two pollutant models with 4-hour ozone 115

Table 2.30. NO2 and day time symptoms in two pollutant models with 4-hour ozone 116

Table 2.31. NO2 and day time medication use in two pollutant models with 4-hour ozone 117

Table 2.32. NO2 and lung function in two pollutant models with 8-hour ozone 118

Table 2.33. NO2 and night symptoms in two pollutant models with 8-hour ozone 119

Table 2.34. NO2 and day time symptoms in two pollutant models with 8-hour ozone 120

Table 2.35. NO2 and day time medication use in two pollutant models with 8-hour ozone 121