Environmental Exposure Assessment Workshop for Population Studies

Banff, AlbertaFeb. 2008

Making choices regarding Making choices regarding characterization of environmental characterization of environmental

exposuresexposures

J.R. BrookEnvironment CanadaUniversity of Toronto

E

Environmental

EExposure

Canadian Healthy InfantLongitudinal Development

The Challenge of Assessing The Challenge of Assessing ExposureExposure

Environmental exposures that have been associated with Environmental exposures that have been associated with the development or exacerbation of childhood asthmathe development or exacerbation of childhood asthma

Fireworks smoke exposure Diet (Vit D, fish consumption)

Ascaris infection, other parasites (protective?) Birth Order/Sibship size

Breastfeeding (mixed) Geography

Outdoor pollen and fungi Household cleaners/pesticides

Early childhood infections Weather/Meteorology

Mould and moisture Housing characteristics

Environmental tobacco smoke Stress

Vaccination Status Endotoxin

Outdoor PM air pollution, traffic pollution Socioeconomic status

Indoor PM air pollution Phthalates, PAHs

House dust mite allergens Pet allergens (cat, dog)

Caesarian section Pest Allergens (cockroach, mouse)

Antibiotic use Chloramines (swimming pools)

Birth weight Violence

Parental asthma/atopy Atopy

Race/Ethnicity Maternal Smoking

There has been decreases in the cost of obtaining detailed genetic information and advances in computational methods.BUTExposure measurement remains expensive and more detail = more expensive Limitations due to ethical considerations

Compromise is necessary

Advances in exposure Advances in exposure characterizationcharacterization

• Measurement technologies– Dust collection approaches– Chemical analysis of dust– Small passive and active samplers– Biomarkers– Greater coverage through satellite observations

• Models– Empirical and physical outdoor air models– Indoor Infiltration models– Probabilistic exposure models– Greater integration

• Questionnaires– More detail (different questions, behaviour)– Ongoing validation

Strategies for prioritizing the Strategies for prioritizing the exposures of interestexposures of interest

• Public Health Importance– exposures that are modifiable, are

common across the population and are hypothesized (or shown) to be associated with the development of asthma

• Measurement Capability– a validated method for characterizing the

exposure is available and resources are available to collect the desired data

• Biological Mechanisms– focus on an exposure that is hypothesized to act

through the same biological pathway as an other exposures that has been associated with asthma

– help develop biomarkers to quantify exposure and/or effect based upon features of the biological pathway

• Potential for Data Pooling– measure the same exposures at the same time

points using the same methods so that the resulting data can be combined with other studies

– requires specific knowledge of the other studies and flexibility in the design of the new study

• larger sample size• potential for greater variability in exposure

Factors Controlling Power inFactors Controlling Power inGene X Environment StudiesGene X Environment Studies

• exposure prevalence– study populations with large exposure prevalence

and contrasts• measurement error

– can be directly influenced by study design– reducing error is an efficient way to improve power

• true effect size– study environmental factors or genetic effects that

are hypothesized to be large based on animal (or microarray) studies

• genotype (phenotype) prevalence– study populations with large prevalence

Wong et al. 2003, IJE

Reducing Measurement ErrorReducing Measurement Error• well-defined research questions• more repeat measurements of exposure• better timing of exposure measurements• use of less expensive exposure surrogates

that are better correlated with the specific outcome of interest

• more accurate/specific genotyping, consistent phenotyping

• improved understanding of the mechanism of effect (epigenetics)– Helps target the desired exposure

Technological AdvancesTechnological Advances

Integrating Sources of Exposure Integrating Sources of Exposure Info.Info.

Measurements + Model + Questionnaires = Measurements + Model + Questionnaires = MODELMODEL

Measurements

Questionnaires

Long

term

tim

e-in

tegr

ated

ex

posu

res

But

at s

peci

fic ti

me

poin

ts (?

)

LUR – Land-Use Regression

Annual Average NOAnnual Average NO22 Observed from Observed from Space Space

TorontoToronto

Best Resolution of Current AQ Best Resolution of Current AQ ModelsModels

Dichotomous Sampler

TEOM

Air Quality Model Measurementnetworks

Empirical models

‘Fused’ spatio-temporalconcentration/exposure

Satellite observations

Hi-res Emissions

Issues to consider when applying Issues to consider when applying LUR to studies of air pollution LUR to studies of air pollution

and healthand health • differential exposures at home,

outdoors and in the workplace• subject mobility• the relevant time period exposure

(and duration of relevant exposure)• exposures to multiple pollutants • tracking effect modifiers• variability in infiltration indoors

Is there a more-direct measure?Is there a more-direct measure?

012345678

0 5 10 15 20

Vehicle exhaust tracer in dust (mg/kg)

R2=0.01

NO2 Outside Home (ppb)

Adj.-NO2 Outside Home (ppb)

012345678

0 5 10 15

R2=0.43Air exchange rate adjusted

detached garageN N

Att. w/door

• Standard biological indicators of allergens• Isolate community DNA for new scans

26.00 28.00 30.00 32.00 34.00 36.00 38.00 40.00 42.00 44.000

10000

20000

30000

40000

50000

60000

70000

80000

90000

100000

110000

120000

130000

140000

150000

Time-->

Abundance

Ion 191.00 (190.70 to 191.70): 9817BM.D

26.00 28.00 30.00 32.00 34.00 36.00 38.00 40.00 42.00 44.000

2000

4000

6000

8000

10000

12000

14000

16000

18000

20000

22000

24000

26000

28000

30000

32000

34000

36000

38000

Time-->

Abundance

Ion 191.00 (190.70 to 191.70): EC100CB.D

Lube o il98-17

PM 98-17US 063

• Chemical Indicators ofmotor vehicle exhaust and phthalate exposures

Combining Existing CohortsCombining Existing CohortsCompromises, compromisesCompromises, compromises

Highest common denominatorHighest common denominator• Outdoor air exposure can often be

reconstructed• Residence at recruitment• Residence during pregnancy and in early life• Complete residence history

– Proximity to source(s)– Dispersion models

• More validation is desired

• Questionnaires – How specific, how validated, how often ?– Geographic/cultural differences in their meaning

• Evaluation of similarities/differences is desired

How do we optimize our How do we optimize our environmental exposure environmental exposure

characterization?characterization?(for G x E research)(for G x E research)

• Many options– Just need time and money

• Efficiency– Comes from asking the right

question(s)

PARSIFAL Schram-Bijkerk et al. 2006 INDOOR AIR

Exposure information that Exposure information that can be collected from can be collected from

questionnairesquestionnaires Housing characteristicsAge, building materials, ventilation systems, below-grade characteristics, renovations, cooking habits, fuel type, swimming pool (chlorine)

Indoor materialsVinyl flooring, carpeting, VOC-based finishes and glues, moisture capturing nutrient rich materials (e.g. dry wall)

MicroenvironmentsTime in traffic, ETS, mother’s occupation, daycare, school ……>> occupation

Animals & PestsCats, dogs, farm animals, rodents, cockroaches, arthopods

Surrounding EnvironmentVegetation (plants, trees, grasses), traffic levels, industrial presence

Consumer products in the homeFurniture, electronics, blinds, cleaning agents, pesticides, herbicides

Dampness and mouldPresence, area of concern, moisture evidence

Exposure information that Exposure information that can be collected from can be collected from

questionnairesquestionnaires Dampness and mould Presence, area of concern, moisture evidence

Activities Time spend indoors, time spent outdoors, time on floor, frequency

of hand-to-mouth contact, storage of toys, sunlight exposureSocioeconomic factors

Income, stress

Diet Food frequency, containers, cooking methods

Infections Viruses, upper respiratory infections

Nitrogen Dioxide vs. Traffic Nitrogen Dioxide vs. Traffic pollutionpollution

200

400

600

800

1000

Hop

anes

(pg

m-3) NO2 (r=0.67 )

0 10 20 30 40 50 60

NO2 (ppb)