11 11 th th EPIET Epidemiology Course EPIET Epidemiology Course Menorca, October 2 2006 Menorca, October 2 2006 Environmental Epidemiology Environmental Epidemiology (introduction) (introduction) Dr Georges Salines Dr Georges Salines Institut de Veille Sanitaire Institut de Veille Sanitaire Département Santé Environnement Département Santé Environnement
66
Embed
11 th EPIET Epidemiology Course Menorca, October 2 2006 Environmental Epidemiology (introduction) Dr Georges Salines Institut de Veille Sanitaire Département.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
► The concept of low-risk and the links The concept of low-risk and the links between relative risk, prevalence of between relative risk, prevalence of exposure and attributable riskexposure and attributable risk
► The limits of epidemiology in environmental The limits of epidemiology in environmental health health
► How to deal with these limitsHow to deal with these limits
DefinitionsDefinitions
► Epidemiology is the study of the distribution Epidemiology is the study of the distribution and determinants of health-related states or and determinants of health-related states or events in specified populationsevents in specified populations
► The environment is all the physical, chemical The environment is all the physical, chemical and biological factors external to a person, and biological factors external to a person, and all the related behaviours. (WHO)and all the related behaviours. (WHO)
► The environment is the sum of all external The environment is the sum of all external conditions affecting the life, development conditions affecting the life, development and survival of an organism (US EPA) and survival of an organism (US EPA)
► The environment is everything that is not me The environment is everything that is not me (Einstein)(Einstein)
►Social factors Social factors (except links between (except links between SES and physical environment)SES and physical environment)
► Infectious diseases Infectious diseases (except those (except those transmitted through exposure to media)transmitted through exposure to media)
RiskRisk
► A measure of the probability that damage to life, health, property, and/or the environment will occur as a result of a given hazard (US EPA)
► Rylander classification RR > = 10 : people themselves recognize the
risk RR de 9 à 2 : « comfort zone » for epidemiology RR < 2 : zone where epidemiology reaches its
limits...
High risksHigh risks
► occupational environmentoccupational environment aromatic amines and bladder canceraromatic amines and bladder cancer asbestos fibres and mesotheliomaasbestos fibres and mesothelioma cadmium and kidney diseasescadmium and kidney diseases benzene and leukaemiabenzene and leukaemia pesticides and infertilitypesticides and infertility organic solvents and neurological disordersorganic solvents and neurological disorders etc ... etc ...
► general environmentgeneral environment
December 1952 - LondonDecember 1952 - London
December 1952 - LondonDecember 1952 - London
1953 - Minamata1953 - Minamata
December 1984 - BhopalDecember 1984 - Bhopal
1986 - Tchernobyl1986 - Tchernobyl
Thyroid cancer in childrenThyroid cancer in children
► mixed originmixed origin UV and melanomaUV and melanoma tremolite and mesothelioma in New Caledoniatremolite and mesothelioma in New Caledonia erionite and mesothelioma in Turkey ...erionite and mesothelioma in Turkey ...
Nature of high risks in general Nature of high risks in general environmentenvironment
► Usually severe and often specific health Usually severe and often specific health endpointsendpoints
► ““well defined” populationswell defined” populations in space, in timein space, in time socio-demographic characteristicssocio-demographic characteristics relatively small populationsrelatively small populations
Characteristics of high risksCharacteristics of high risks
► urban air pollution and short-term urban air pollution and short-term respiratory diseasesrespiratory diseases RR = 1.1 - 1.5RR = 1.1 - 1.5
► chlorinated water supplies and bladder chlorinated water supplies and bladder cancercancer RR = 1.4RR = 1.4
► electromagnetic fields and children electromagnetic fields and children leukemialeukemia RR = 1.3RR = 1.3 ... ...
LowLow risksrisks
►Relative risk and attributable riskRelative risk and attributable risk relative riskrelative risk
►ratio measure : “ratio measure : “it is an indicator for it is an indicator for epidemiologist ”epidemiologist ”
attributable riskattributable risk►FRA = p FRA = p ** ( RR -1) / [ 1+ p ( RR -1) / [ 1+ p ** ( RR - 1) ] ( RR - 1) ]
““if the relation is causalif the relation is causal, it estimates the , it estimates the proportion (amount) of diseases that we proportion (amount) of diseases that we can attribute to the exposure”can attribute to the exposure”
Small relative risks do not mean Small relative risks do not mean small health impactssmall health impacts
Exposure prevalence
Relative
Risks
0,01 0,10 1,00
10 0,08 0,47 0,90
5 0,04 0,29 0,80
2 0,01 0,09 0,50
1,5 - 0,05 0,33
1,2 - 0,02 0,17
1,1 - 0,01 0,09
Health impact
Exposure prevalence
Relative
Risks
0,01 0,10 1,00
10 0,08 0,47 0,90
5 0,04 0,29 0,80
2 0,01 0,09 0,50
1,5 - 0,05 0,33
1,2 - 0,02 0,17
1,1 - 0,01 0,09
Health impactHealth impact
May be not May be not thatthat low after all low after all
E0 = non exposed, E1=low exposure, E2=high exposureE0 = non exposed, E1=low exposure, E2=high exposure* Incidence : x /100.000, * Incidence : x /100.000, ** RR : true Relative Risk** RR : true Relative Risk
Heterogeneity in the population’s Heterogeneity in the population’s sensitivitysensitivity
to the exposureto the exposure E0 E1 E2
Prevalence 80% 15% 5%
Incidence (S) 100 300 500
Incidence (s) 100 200 300
100 250 400
RR 1.0 2.5 4.0
50%50%
50%50%
* (S) : high sensitivity. (s) : low sensitivity
Non specific definition of the health Non specific definition of the health outcomeoutcome
E0 E1 E2
Prevalence 80% 15% 5%
Incidence (D) 100 250 400
Incidence (d) 50 50 50
150 300 450
RR 1.0 2.0 3.0
* (D) : disease specifically related to exposure. (d) : disease not related to exposure
Errors in the exposure classificationErrors in the exposure classification
20% of non exposed (E0) are categorised E1 and 10% of non-exposedare categorised E2.
E0 E1 E2
Prevalence 50% 35% 15%
Incidence 150 214.3 250
RR 1.0 1.43 1.67
Inaccuracy in the exposure Inaccuracy in the exposure categoriescategories
E0 E1
Prevalence 50% 50%
Incidence 150 225
RR 1.0 1.5
Epidemiology and weak Epidemiology and weak associationsassociations
► Improve data qualityImprove data quality exposureexposure health endpointshealth endpoints co-factorsco-factors
► Improve statistical powerImprove statistical power Meta-analysis & Multi Meta-analysis & Multi
centrescentres
►Ecological designsEcological designs
appropriate selection of sources and appropriate selection of sources and routes of exposureroutes of exposure
taking account:taking account:►critical periods of exposurecritical periods of exposure►individual history of exposure : behaviour, individual history of exposure : behaviour,
space-time activities …space-time activities …
Improving assessment of exposure:Improving assessment of exposure:better use of environmental databetter use of environmental data
Exposure = last place of residence, assuming entire life at this residence and consumption of the town’s drinking water Chlorination exposure
0 1-25 26-50 50+
Cases (286)
84
31
79
92
Controls (658)
193
104
188
173
OR 1.0 0.7 1.0 1.2
IC 95% 0.4 - 1.1 0.7 - 1.4 0.9 - 1.8
Lynch et al, Arch Env Health 1989;44(4):252-259Lynch et al, Arch Env Health 1989;44(4):252-259
ExampleExample
Exposure = entire lifetime residential history combined with the town’s drinking water chlorination history
Chlorination exposure
0 1-25 26-50 50+
Cases (286)
65
92
102
27
Controls (658)
211
211
195
41
OR 1.0 1.4 1.7* 2.1*
IC 95% 1.1 - 2.1 1.2 - 2.5 1.2 - 3.7
Lynch et al, Arch Env Health 1989;44(4):252-259Lynch et al, Arch Env Health 1989;44(4):252-259
Example (2)Example (2)
technical, logistical and financial limits …technical, logistical and financial limits … depends on sensibility / specificity of the depends on sensibility / specificity of the
methodmethod
Improving assessment of exposure:Improving assessment of exposure:personal exposure monitoringpersonal exposure monitoring
cellular, biochemical, molecular alterationscellular, biochemical, molecular alterations► measurable in biological media (human tissues, cells or measurable in biological media (human tissues, cells or
fluids)fluids) advantages advantages
► measurement of a dose (effectively absorbed)measurement of a dose (effectively absorbed)► integration of all the routes of exposure and sources of integration of all the routes of exposure and sources of
absorptionabsorption► avoids subjects’ lack of knowledge, memory failure, avoids subjects’ lack of knowledge, memory failure,
►Example 2: Flight of the sick people (Bias toward false negative)
Ecological Fallacy in Geographical study
Environmental exposure
Area B
Area C
Area A
Incidence rate
Ecological Fallacy
population B
population C
population A
Individual exposure
Incidence rate
ExampleExample
►1983: leukaemia cluster among 1983: leukaemia cluster among children living near the Sellafield children living near the Sellafield nuclear waste reprocessing plant nuclear waste reprocessing plant (United Kingdom)(United Kingdom)
►Other leukaemia clusters have since Other leukaemia clusters have since been identified near other nuclear been identified near other nuclear sites, such as Dounreay in Scotland sites, such as Dounreay in Scotland and Krümmel in Germanyand Krümmel in Germany
But…But…
► In view of current knowledge about the relation In view of current knowledge about the relation between exposure to radiation and the risk of between exposure to radiation and the risk of leukemia, dose levels around nuclear sites are leukemia, dose levels around nuclear sites are incompatible with the excess risks observed …incompatible with the excess risks observed …
► Studies considering several sites (United Kingdom, Studies considering several sites (United Kingdom, France, USA, Germany, Canada, Japan, Sweden, France, USA, Germany, Canada, Japan, Sweden, Spain) have not detected any global excess Spain) have not detected any global excess
► Leukaemia clusters have been observed in areas Leukaemia clusters have been observed in areas far from any nuclear sitefar from any nuclear site
► There are alternative hypotheses which may There are alternative hypotheses which may explain the leukaemia clusters located near some explain the leukaemia clusters located near some nuclear sitesnuclear sites
Interpretation of geographical studies
►Measures of geographical associations
►Very difficult to extrapolate at the individual level
►Causality generaly out of reach of those designs
►Useful for generating hypotheses
Time seriesTime series
►Statistical powerStatistical power►Control of confounding factors Control of confounding factors
++++++ Non time-dependant: Population is its Non time-dependant: Population is its
own controlown control Time-dependant: modelling Time-dependant: modelling
techniques techniques
Exemple PSAS9 IExemple PSAS9 I
D day
Exposed population
Indicator of exposure
Indicatorof effect
All people living in
Marseilles
SO2 g/m3
(Daily mean of 3 monitoring stations)
Daily number of deaths
Source : Surveillance épidémiologique air et santé, rapport InVS, mars 1999
Raw curves
Dioxyde de soufre (SO2)
µg
/m3
02
04
06
08
01
00
Jan 1990 Jan 1991 Jan 1992 Jan 1993 Jan 1994 Jan 1995
Fumées noires (FN)
µg
/m3
02
04
06
08
0
Jan 1990 Jan 1991 Jan 1992 Jan 1993 Jan 1994 Jan 1995
Mean levels of air pollution Marseilles, 1990-1995
5
10
15
20
25
30
35
40
Daily counts of deaths, Marseilles, 1990-1995
Source : Surveillance épidémiologique air et santé, rapport InVS, mars 1999
Time-dependant counfounding factors
Jour
PA
CF
0 5 10 15 20 25 30
-0.1
0-0
.05
0.0
0.05
0.10
0.15
0.20
Fonction totale.
Source : Surveillance épidémiologique air et santé, rapport InVS, mars 1999
Serial correlation fonction of daily mortality
Time-dependant counfounding factors
Risk factor Health outcome
Confounding factor
Daily variation of air pollution levels
Daily variation of number of deaths from respiratory causes
Daily variations of weather
Filleul et coll., Rev. Mal. Respir., 2001
Non time-dependant counfounding factors
Risk factor Health outcome
Confounding factor
Daily variation of air pollution levels
Daily variation of number of deaths from respiratory causes
Tobacco
Filleul et coll., Rev. Mal. Respir., 2001
Modeling: Strip-tease of the curves
► Taking into acount long-term trends Taking into acount long-term trends (ie:decrease of mortality)(ie:decrease of mortality)
► Taking into acount seasonal variations Taking into acount seasonal variations (Higher mortality during winter)(Higher mortality during winter)
► Taking into acount the day of the weekTaking into acount the day of the week► Taking into acount co-factors Taking into acount co-factors
Predicted value of total mortality by trend-modeling.
Source : Surveillance épidémiologique air et santé, rapport InVS, mars 1999
21
21.5
22
22.5
23
23.5
24
Seasonal variations
Predicted value of mortality by modelization of seasonal variations
Source : Surveillance épidémiologique air et santé, rapport InVS, mars 1999
19
21
23
25
27
29
31
Meteorological data
Naperian Logarithm of Relative Risk of the interaction temperature-humidity on total mortality
Source : Surveillance épidémiologique air et santé, rapport InVS, mars 1999
Day of the week
Predicted value of total mortality by modelization of a « day of the week & holidays» effect
Source : Surveillance épidémiologique air et santé, rapport InVS, mars 1999
15
17
19
21
23
25
27
29
31
33
Full Monty
-4
-3
-2
-1
0
1
2
3
4
Jour
PA
CF
0 5 10 15 20 25 30
-0.1
0-0
.05
0.0
0.0
50
.10
0.1
50
.20
Serial correlation fonction of daily mortality after modelization of trend, seasonal variatipons, Flu epidemics, temperature, humidity, day of the week & holidays
Residual values of total mortality after modelization of trend, seasonal variatipons, Flu epidemics, temperature, humidity, day of the week & holidays
Source : Surveillance épidémiologique air et santé, rapport InVS, mars 1999
Result “dose-response curve” of total mortality in relation to SO2 levels
Source : Surveillance épidémiologique air et santé, rapport InVS, mars 1999
SO2 moyenne 0-1 jours
RisqueRelatif
0 20 40 60 80
1.00
1.02
1.04
1.06
1.08
1.10
non-paramétrique
linéaire
Interpretation of time-series Interpretation of time-series studiesstudies
► Establishing causation Establishing causation is possibleis possible after a after a careful discussion of Hill criteriacareful discussion of Hill criteria1.1. Strength. Strength. 2.2. Consistency. Consistency. 3.3. Specificity. Specificity. 4.4. Temporality. Temporality. 5.5. Biological gradient (dose-response). Biological gradient (dose-response). 6.6. Plausibility. Plausibility. 7.7. Coherence. Coherence. 8.8. Experiment. Experiment. 9.9. Analogy. Analogy.
V. ConclusionV. Conclusion
► Aspects of the study design that involves Aspects of the study design that involves measurements of variables are critical, especially measurements of variables are critical, especially in environmental epidemiology where risks from in environmental epidemiology where risks from exposure are likely to be small, difficult to detect, exposure are likely to be small, difficult to detect, and perhaps not clinically significant, yet maybe and perhaps not clinically significant, yet maybe of public health importanceof public health importance
► Epidemiology is not always the only answer of Epidemiology is not always the only answer of even the more relevant one to questions even the more relevant one to questions submitted to environmental epidemiologists: Risk submitted to environmental epidemiologists: Risk analysis for example, is a very useful and cost-analysis for example, is a very useful and cost-effective methodeffective method
► ...but this is another story....but this is another story.