Meeting of the Eionet National Reference Centres for Noise Copenhagen, 14 th - 15 th October 2009 Linking Exposure To Health Wolfgang Babisch Martin van.

Meeting of the Eionet National Reference Centres for NoiseCopenhagen, 14th - 15th October 2009

Linking Exposure To Health

Wolfgang Babisch

Martin van den Berg

European Environment AgencyExpert Panel on Noise (EPoN)

European Environment Information and Observation Network

Calm NetworkVision for the Year 2020

Source: CALM, 2004

“Avoid harmful effects of noise exposure from all sources and preserve quiet areas.”

Community Research Strategy Plan

Coordination of European Research for Advanced Transport Noise Mitigation

WHO Constitution (1948):

Health is a state of complete physical mental and social well-being

and not merely the absence of disease or infirmity.

Health

Sound level

Annoyance

Noise sensitivity

Other factors

Health

Sound level

Annoyance

Noise sensitivity

Other factors

Health

Well-being and Ill-health

Introduction

Health endpoints

Exposure-response curves

Risk assessment

Cost-benefit analysis

Implications for END

References

Expert Panel on Noise (EPoN)Fact Sheet

Fact-sheet on Noise Exposure andPotential Health Impact

Introduction

Health endpoints

Exposure-response curves

Risk assessment

Cost-benefit analysis

Implications for END

References

Expert Panel on Noise (EPoN)Fact Sheet

Fact-sheet on Noise Exposure andPotential Health Impact

Hazard identification

Exposure-response assessment

Exposure assessment

Risk characterizationAttributable risk concept

Risk managementRegulatory options

Quantitative Risk Assessment

END ?Link

Research !

Hearing loss

Annoyance

Sleep disturbance (objective and subjective)

Cognitive impairment

Mental health

Physiological stress reaction

Cardiovascular risk

Health Endpoints"Hazard Identification"

Source: Miedema and Oudshoorn, 2001European Commission Working Group on Dose-Effect Relations, 2002

0

10

20

30

40

50

60

70

80

35 40 45 50 55 60 65 70 75 80

Noise level (Lden, 24h) [dBA]

Pe

rce

nt

hig

hly

an

no

ye

d (

HA

) [%

] Aircraft

Road

Railway

25%-Criterion

Common rule

WHO Community Noise Guidelines (1999):

"Serious annoyance"

Annoyance "Highly Annoyed"

Aircraft noise

0

10

20

30

40

50

60

70

80

40 45 50 55 60 65 70 75

Lden (aircraft noise) [dB(A)]

Pe

rce

nt

hig

hly

an

no

ye

d [

%]

Miedema Total GB D NL S

Δ = 7-8 dB(A)

0

10

20

30

40

50

60

70

80

40 45 50 55 60 65 70 75

Lden (road traffic noise) [dB(A)]

Pe

rce

nt

hig

hly

an

no

ye

d [

%]

Miedema Total GB D NL S

Road traffic noise

Change of Annoyance "Hyena Study"

Source: Babisch et al., 2009

Annoyance Due To Aircraft NoiseUpdate

"Little annoyed" "Highly annoyed""Annoyed"

Source: TNO, 2009

Older studies

Older studies

Older studies

Newer studiesincluded

Newer studiesincluded

Newer studiesincluded

Sleep

Normal sleep

Noise disturbed sleep

Source: Maschke et al., 2000

Sleep Disturbance (Objective)Awakenings (Single Aircraft Flyover) – "DLR Laboratory Study"

(Changes to wake or to sleep-stage S1 within 90 seconds)

Maximum Sound Pressure Level LAS,max [dB]

Source: Basner et al., 2006

Pro

babi

lity

of s

leep

sta

ge c

hang

es to

aw

ake

Spontaneous awakening

Noise induced awakening

33 dB

Sleep Disturbance (Objective)Aircraft Noise Simulation – "DLR Field Study"

Source: Basner and Griefahn, 2008

Moderate Effects in

Vulnerable Groups

Steep Increase of

Adverse Effects

Public Health Hazard

Lnight, outside (aircraft)Figure 1: The average number of additionally aircraft noise induced awakenings per night. Altogether, 10 million 8-hour nights with 1 to 200 (1, 2, 3,…, 200) noise events randomly drawn from the DLR field study 20 were simulated. The lines represent (from below to above) 2.5, 25, 50, 75, and 97.5 percentiles. In an 8 h undisturbed night of a healthy sleeper, on average 24 spontaneous EEG awakenings (as defined by Rechtschaffen and Kales1) can be observed (equaling 8,760 spontaneous awakenings per year).

365 EEG awakenings per year

• Noise and transient sleep disturbance - well developed area with statistically robust data and dose-response relationships.

• No quantitative link yet established between acute or transient sleep disturbance caused by noise and any long term adverse health effects.

• No consensus on any single dose-response relationship which could be used to inform cost benefit analysis, monetary evaluation of adverse health effects, policy etc.

Sleep Disturbance Health (?)

Source: Bell report, 2009Griefahn and Basner (Internoise), 2009

Cross-sectional study, 1718 males and females, aged 18-90 yrsRelative prevalence of treated high blood pressure by road traffic noise level

0

1

2

3

4

< 60 60 - 65 > 65

Lday,16h [dB(A)]

Od

ds

Rat

io

0

1

2

3

4

< 50 50 - 55 > 55

Lnight,8h [dB(A)]

Od

ds

Rat

io

Significant effects only with respect to the exposure of the bedroom

OR = 1.9

Day - Living room Night - Bedroom

Source: UBA - Maschke et al. (2003)

Hypertension"Spandau Health Survey"

0

5

10

15

20

25

30

35

35 40 45 50 55 60 65 70 75 80Noise level (Lnight) [dBA]

Per

cen

t h

igh

ly s

leep

dis

turb

ed (

HS

D)

[%]

Aircraft

Road

Railway

5%-Criterion (?)

WHO Community Noise Guidelines (1999):

"Sleep disturbance, window open"

Source: Miedema and Oudshoorn, 2001European Commission Working Group on Health and Socio-Economic Aspects, 2004

Sleep Disturbance (Subjective) "Highly Sleep Disturbed"

WHO Night Noise Guidelines For Europe (2009):

"Night noise guideline"

Source: Hygge et al., 2002

no noise

noise

Difficult word list, 326 children

Old airport

Wave 1 Wave 2 Wave 3

Old airport closed

0

1

2

3

4

5

6

7

8

9

10

Me

an

err

ors

Mea

n er

rors

New airport

0

1

2

3

4

5

6

7

8

9

10

Me

an

err

ors

Wave 1 Wave 2 Wave 3

New airport opened

Mea

n er

rors

Cognitive Impairment"Munich Study"

Source: Stansfeld et al., 2005

-.2

0.2

.4R

eadi

ng Z

-sco

re

30 35 40 45 50 55 60 65 70aircraft noise dB(A)

2844 children

Reading Comprehension"Ranch Study"

Aircraft noise at school LAeq,16h [dB]

Aircraft noise associated with deficits in reading comprehension and recognition memoryR

eadi

ng Z

-sco

re

NoiseWorking memory

Episodic memory

Read

Listen

Phonological loop

Visual loop

Basic Memory ModelRead – Listen - Learn

It could be noted that:

Both chronic noise exposure in the field study and acute noise exposure in the experiments impair episodic memory, in particular recall.

There is a strong similarity in how much and in what way children and adults are affected by noise.

Staffan Hygge

Source: Hygge, 2009 (presentation in Helsinki)

Ldn

0

2 0

4 0

6 0

8 0

1 0 0

4 0 4 5 5 0 5 5 6 0 6 5 7 0 7 5 8 0 8 5 9 0 9 5 1 0 0

L d n

Per

cen

t af

fect

ed

L d n

Assuming that 100% of the noise exposed are cognitively affected at a very high noise level, e.g. 95 Ldn, and that none are affected at a safely low level, e.g. 50 Ldn, a straight line (linear

accumulation) connecting these two points, as in the Figure below can be used as basis for approximation.

Hypothetical Exposure-Response Curve- Cognitively Affected -

Source: Hygge, 2009 (presentation in Helsinki)

Staffan Hygge

No evidence

Mental Health

v. Kamp and Davies (2008):

"... there is no direct association between environmental noise and mental health". "Noise annoyance is consistently found to be an important mediator."

Clark and Stansfeld (2007):

"Overall, studies suggest that both adults and children noise exposure is probably not associated with serious psychological illness, but there may be effects on well-being and quality of life."

Source: Bell report, 2009

Source: Henry & Stephens, 1977PerceivedStimulus

Coping PatternsEarly Experience

Genetics

Fight - Flight

Nervous System + Adrenaline + Noradrenaline o Cortisol

Hypophysis (Pituitary Gland) + ACTH + Cortisol o Catecholamines

Depression

Threat to Control Loss of Control

Sympathetic-Adrenal-Medullar Axis Hypophysio-Pituitary-Adrenal-Cortical Axis

Stress Model

Cross-sectional study, 195 females, aged 30-45 yrsOvernight excretion of norepinephrine in urine by road traffic volume

Mean +/- SE

20

21

22

23

24

25

26

<14,000 (mean = 2,300) >20,000 (mean = 31,000)

Traffic volume [vehicles/day]

Noradrenalineper Creatinine

[µg/g]

Source: Babisch et al. (2001)

LAeq,24h < 60 dB LAeq,24h > 65 dB

Stress Hormones"Berlin Traffic and Health Study"

- With respect to environmental noise acute noise effects occur, in particular, when certain activities such as concentration, relaxation or sleep are disturbed.

- Even subjects that have been living for many years in exposed homes show physiological stress reactions.

- During sleep no complete habituation takes place. Even subjects who are subjectively disturbed by the noise show acute physiological reactions to single noise events.

Figure: Maschke, 2004

Habituation / Adaption

Sound exposure

Disturbance ofintended activities

Stress indicators

Biologicalrisk factors

Cardiovasculardiseases

Noise Exposure (Sound Level)

Direct pathway Indirect pathway

Hearing loss

Disturbance of activities, sleep communicationCognitive and

emotional response

Annoy-ance

Autonomic nervous system (sympathetic nerve)Endocrine system (pituitary gland, adrenal gland)

Cardiovascular DiseasesHypertension Arteriosclerosis Ischaemic heart diseases

Stress Indicators

Risk Factors

Blood pressure Blood lipids Blood viscosityCardiac output Blood glucose Blood clotting factors

Manifest Disorders

Physiological stress reactions (homeostasis)

Simplified Noise Reaction Model

Source: Babisch, 2002

-0.2

-0.1

0

0.1

0.2

0.3

0.4

51-55

56-60

61-65

66-70

51-55

56-60

61-65

66-70

Traffic Noise Level [dB(A)]

Re

lati

ve

ch

ole

ste

rol l

ev

el

[mm

ol/l

]

+ SE

Mean

- SE

Caerphilly Speedwell

Cross-sectional study, 4860 males, aged 45-63 yrsRelative concentration of serum cholesterol by road traffic noise

Source: Babisch et al., 1993

Traffic noise level Lday,16h [dB(A)]

Serum Cholesterol"Caerphilly & Speedwell Studies"

Cross-sectional study, 4860 males, aged 45-63 yrsRelative concentration of plasma viscosity by road traffic noise

-0.01

-0.005

0

0.005

0.01

0.015

0.02

0.025

0.03

51-55

56-60

61-65

66-70

51-55

56-60

61-65

66-70

Traffic Noise Level [dB(A)]

Re

lati

ve

pla

sm

a v

isc

os

ity

lev

el

[cp

] + SE

Mean

- SE

Caerphilly Speedwell

Traffic noise level Lday,16h [dB(A)]

Source: Babisch et al., 1993

Plasma Viscosity"Caerphilly & Speedwell Studies"

Source: Babisch, 2006

0.6

0.8

1

1.2

1.4

1.6

1.8

Lday [dB(A)]

Od

ds

rati

o

Caer+Speed

Berlin I

Berlin II

Berlin III

Pooled

Traffic noise level Lday,12hr [dB(A)]

Heterogeneity: p = 0.943

Exposure-Response Curve: Myocardial InfarctionMeta-Analysis: Road Traffic Noise

5 studies

Source: Babisch, 2006, 2008

Exposure-Response Curve: Myocardial InfarctionMeta-Analysis: Road Traffic Noise

Traffic noise level Lday,12hr [dB(A)]

Re

lati

ve

ris

k (

od

ds

ra

tio

)

Ischaemic Heart Disease

Exposure-response function: OR = 1.629657 – 0.000613*(Lday,16h)2 + 0.000007357*(Lday16h)3, R2 = 0.96OR per 10 dB(A) = 1.17, 95% CI = 0.87-1.57, range = 60-80 dB(A)

pooled(5 studies)

Source: Babisch und van Kamp, 2008

Exposure-response function: OR per 10 dB(A) = 1.13, 95% CI = 1.00-1.28, range = 50-70 dB(A)

Exposure-Response Curve: HypertensionMeta-Analysis: Aircraft Noise

0.8

1

1.2

1.4

1.6

1.8

2

2.2

45 50 55 60 65 70 75

Aircraft noise level approx. Ldn [dB(A)]

Re

lati

ve

ris

k Amsterdam

Stockholm1

Okinawa

Hyena

Stockholm2

pooled(5 studies)

Heterogeneity: p = 0.002

Hyena study, 4861 subjects, age 45-70 yrsRelative prevalence of hypertension by road traffic noise

0.190 0.928 0.829-1.038

0.031 1.141 1.012-1.286

P = 0.044 OR pro 10 dB(A) = 1.097 CI = 1.003-1.201

0.550 1.021 0.953-1.095

0.037 1.101 1.006-1.205

0.068 1.071 0.995-1.154

0.041 1.099 1.004-1.202

0.8

0.9

1

1.1

1.2

1.3

1.4

1.5

1.6

1.7

1.8

<47.5 47.5-52.4 52.5-57.4 57.5-62.4 62.5-67.4 >=67.5

LAeq24h - road [dB(A)]

Od

ds r

ati

o +

/- 9

5%

CI

Source: Jarup et al., 2008

Road Traffic Noise - Hypertension"Hyena Study"

Hazard identification

Exposure-response assessment

Exposure assessment

Risk characterizationAttributable risk concept

Risk managementRegulatory options

Quantitative Risk Assessment

Link

Annoyance

Sleep disturbance (subjective)

Cognitive impairment (children)

Cardiovascular

Established Exposure Response Relationships

Population attributable risk

Impact fraction

AF = {(Pi * RRi) - 1} / (Pi * RRi)

where: Pi = Proportion of the population in exposure category i RRi = relative risk at exposure category i compared to the reference level

Attributable FractionFormula

7-19 year old children

Noise exposure level Ldn

No. of children exposed *

% who will develop noise-induced cognitive impairment

N with noise-induced cognitive impairment per year

< 55 68% 1 012 817 0 0

55-65 19% 282 993 20 56 599

65-75 11% 163 838 50 81 919

> 75 2% 29 789 75 22 342

Total 100% 1 489 437 160 859

Attributable Fraction: Cognitive ImpairmentExample: Sweden, Children

Source: Hygge, 2009 (presentation in Helsinki)

Staffan Hygge

* Approximation: European exposure distribution considered for Sweden (Roovers et al., 2000)

Moyocardial Infarction (MI) Caused By Traffic Noise

Reference year 1999

Average sound pressure level

during the day (6-22 h)

Lday,16hr [dB(A)]*

Percentage exposed

[%]

Relative risk of myocardial infarction

[OR] *)

<= 60 69.1 1.000

>60 – 65 15.3 1.031

>65 – 70 9.0 1.099

>70 – 75 5.1 1.211

>75 1.5 1.372

Attributable fraction: 2.9%

Germany 1999:

MI: 133,115 cases

IHD: 849,557 cases

Attributable Fraction: Myocardial Infarction / IHDExample: Germany, Adults

* Probabilistic model

Due to road traffic noise:

3,900 MI cases/year

24,700 IHD cases/year

MI = Myocardial infarctionIHD = Ischaemic heart diseases

Source: Babisch, 2002 (WHO 1992 - modified)

Severity vs. Population Affected

Mortality

Disease(sleep disturbance,

cardiovascular)

Stress Indicators(autonomous response, stress hormones)

Risk factors(blood pressure, cholesterol,

blood clotting, glucose)

Feelings of discomfort(annoyance, disturbance)

Number of people affected

Source: de Hollander et al., 2001, 2004

Life expectancy: mortality, morbidity, loss of healthy life-years    

Quality of life: severe feelings of discomfort, reduced ability to concentrate, unfavourable health perception and stress in relation to poor quality of the local environment

Social magnitude: number of people affected

DALYsDisability Adjusted Life Years

Idea: Quantification UnitOne figure to aggregate different effects of different severity

Priority settingDALYs as a public health currency?

Cost-benefit analysisMonetary valuation of external costs and benefits (DALYs vs. QUALYs)

DALY = YLL x YLDYLL = ND (number of deaths) x DW (disability weight) x LD (standard life expectancy at age of death in years)YLD = NI (number of incident cases) x DW (disability weight) x LI (average duration of disability in years)

Severity weight factors (disability weights)Ethics (dependent on cultural weighing systems), who decides (experts)?

DALYs

Mortality = 1.000

Non-fatal acute myocardial infarction = 0.406 (WHO)

Ischaemic heart disease = 0.350 (de Hollander, 1999)

High blood pressure = 0.352 (Mathers, 1999)

Primary insomnia = 0.100 (WHO, 2007)

Sleep disturbance = 0.020 (Knol, 2005); 0.055 (Müller-Wenk, 2005)

Annoyance = 0.010 (Stassen, 2008); 0.033 (Müller-Wenk, 2005)

Cognitive impairment = 0.006 (Hygge, 2009)

Disability WeightsDALYs

Disability weights that have been used

Noise exposure levelLden

No. exposed population

< 40 3 %

40-50 17 %

50-60 56 %

> 60 24 %

Total 100 %

Attributable Fraction: AnnoyanceExample: DALYs Per Year For The Netherlands

Source: Knol and Staatsen, 2005

Reference year 2000

Exposure distribution ('Empara')

Noise source DALYs per million people per reference year

Road 1122

Air 16

Rail 65

Total 1203

"Severe" annoyance('Miedema' curves)

Adults:Disability weight DW = 0.020

Example:Severe

annoyanceNetherlands

Severe sleep disturbance Netherlands

Cognitive effectsSweden

Ischaemic heart diseasesGermany

Subjects Adults Adults Children 7-19 yrs Adults

Total population Ca. 14 Mio Ca. 14 Mio Ca. 1.5 Mio Ca. 70 Mio

Exposure Empara Empara EU estimate Probabilistic UBA

Reference year 2000 2000 2000 1999

Disability weight 0.02 0.02 0.006 0.350

DALYs per Million of people

1203 591 648 361

DALY Examples

Economic ValuationExample: Switzerland

Source: Swiss Federal Office for Spatial Development, 2004, 2008

Exposure Assessment

Environmental Noise Directive, Noise Mapping

Which purpose?

- Identification of hot spots (exposure)

- Public health (affected population in general)

- Action plans

- Regulatory (limit values)

- Epidemiological study (individual exposure)

Shielding: Terraced HousesLday,12 h

Distance from the road (meters)Source: Lercher et al. (2000)

Exposed Side: High Blood Pressure"Inn Valley Study"

Proximity to road, room orientation

Pro

po

rtio

n w

ith

hy

per

ten

sio

n

Source: Lercher et al, 2000

Modifiers of Exposure: Windows"Inn Valley Study"

Systolic blood pressure

Prospective study, 3950 men, aged 45-63 yrsExtreme group comparison: LAeq,day = 66-70 dB(A) vs. 51-55 dB(A)

Relative incidence of major ischaemic heart diseases

Source: Babisch et al. , 1999

0

0.25

0.5

0.75

1

1.25

1.5

1.75

2

2.25

2.5

Addressonly

+ Windoworientation

+ Openwindow

+ 15 yrsresidence

Inc

ide

nce

od

ds

rati

o(O

R +

/- 9

5% C

I)

Exposure Misclassification"Caerphilly & Speedwell Studies"

Environmental Noise Directive(Directive 2002/49/EG)

Phase 1 until 30.06.2007» Major road network > 6 Mio vehicles per year» Major railway network > 60.000 trains per year» Major airports > 50.000 aircraft movements per year» Major agglomerations > 250,000 inhabitants» Action plans until 2008

Phase 2 until 30.06.2012» Major roads > 3 Mio vehicles per year» Major railway tracks > 30.000 trains per year» Agglomerations > 100,000 inhabitants» Action plans until 2013

Augsburg Noise Maps House At A Close Distance To A Major Road

City Map ("RLS90") END Map ("VBUS")

Software: Cadna

Augsburg Noise MapsHouse At A Close Distance To A Major Road

City Map

Software: Cadna

Front / back sideLday, 16h: 70/50 dB(A)Δ = 20 dB(A)

END Map

Front / back sideLday, 12h: 67/44 dB(A)Δ = 23 dB(A)

Augsburg Noise MapsHouse At A Far Distance To A Major Road

City Map END Map

Software: Cadna

Augsburg Noise MapsHouse At A Far Distance To A Major Road

City Map

Software: Cadna

Front / back sideLday, 16h: 62/46 dB(A)Δ = 16 dB(A)

END Map

Front / back sideLday, 12h: 44/43 dB(A)Δ = 1 dB(A)

Exposure-response relationships are available

Methods for quantitative risk assessment are established

Applicable for general noise policies and action planning ("what happens - if" scenarios)

DALYs for combining effects (?)

Economic valuation for decision making (?)

Exposure assessment must be comprehensive

END underestimates the true exposure

Conclusions

Contact:

[email protected]

“Calling noise a nuisance is like calling smog an inconvenience”(W. H. Steward, former Surgeon General of USA)

Thank You For Listening

Cardiovascular: Lday ≤ 65 dB(A)

Serious annoyance: Lday ≤ 55 dB(A)

Moderate annoyance: Lday ≤ 50 dB(A)

Cardiovascular: Lnight ≤ 55 dB(A)

Sleep disturbance: Lnight ≤ 45/40 dB(A)

Outdoors

Environmental Noise Quality Targets

» Short-termPrevention of health riskLden = 65, Lnight = 55

» IntermediateReduction of serious annoyance Lden = 60, Lnight = 50

» Long-term Avoidance of serious annoyance Lden = 55, Lnight = 45

UBA Quality Targets"Federal Environment Agency, Germany"

Coherence (biological plausibility)

Consistency with other studies in different populations and with different methodology

Presence of dose-response relationship

Magnitude of effect

Even small relative risks may be relevant for public health

Epidemiological Reasoning

Cross-sectional study, 195 females, aged 30-45 yrsOvernight excretion of nor-epinephrine in urine by road traffic volume

Source: Babisch et al., 2001

20

21

22

23

24

25

26

27

28

29

30

Low (categories 1+2+3) High (categories 4+5)

Sleep disturbance

No

rad

ren

aln

e [µ

g/g

Cre

atin

ine]

Windows open

Windows closed

Subjects had controlover the stimulus

Subjects had no control over the noise

Coping: Stress Hormones"Berlin Traffic And Health Study"