Selected exposures and health effects

This chapter presents, for each stressor, the health effects included in our analyses, exposure response functions that were used and the exposure data. Table 3-19 presents an overview of the selected environmental stressors, health endpoints, exposure-response functions and methods used. An overview of exposure data used for each stressor is given in Table 3-20.

Selection criteria

Environmental stressors

We aimed to study the burden of disease in the general population associated with stressors in the physical environment. Occupational hazards and risks associated with lifestyles (e.g. alcohol use, active smoking, nutrition), as well as infectious diseases, were excluded from the assessment.

Four criteria were defined for selection of environmental stressors to be included in the study:

• Public health impact;
• High individual risk;
• High political or public concern;
• Economic significance.

High priority list of environmental stressors
Benzene
Dioxins (including furans and dioxin like PCBs)
Second-hand smoke (SHS)
Formaldehyde
Lead
Transport noise
Ozone
Particulate matter
Radon

Medium priority list of environmental stressors
1,2-Dichloroethane
Accidents - domestic
Accidents - traffic
Acrylamide
Arsenic
Chlorination by-products
Carbon monoxide (CO)
Damp housing
Foodborn epidemics
Indoor insecticides
Methyl mercury
UV radiation
Waterborne epidemics

In addition, the selection was affected by the feasibility of the calculation. Therefore, we also considered:

• availability of exposure data
• availability of evidence-based exposure response function(s)
• availability of baseline health statistics.

Discussion among environmental health experts as represented in the EBoDE working group selected the environmental stressors based on these selection criteria. A first list of stressors was divided into two parts:

• a high priority list of stressors, which either scored high on many of the criteria and/or which were relatively easy to calculate
• a medium priority list of stressors.

The pilot project, which is described in this report, only included the stressors on the high priority list. These stressors will be shortly introduced in paragraphs 3.2 to 3.10. This list represents only a limited number of environmental stressors, and therefore the results of this study cannot be interpreted as estimates of the complete environmental portion of the total burden of disease.

The stressors on the medium priority list are candidates for addressing in subsequent studies.

Health outcomes

For every environmental factor, a set of health endpoints had to be selected which are causally linked to the exposure of interest. Only health effects that are included in the International Statistical Classification of Diseases and Related Health Problems (ICD) were selected. Therefore, wellbeing effects and for example ‘noise annoyance’ were not included. Within that definition of health, the health endpoints were selected based on the following criteria:

• “sufficient” evidence for a causal relationship between exposure to the environmental stressor and the health effect
• “sufficient” evidence that the health effect is substantive enough to have an impact on the burden of disease estimate
• sufficient data to carry out the calculations (burden of disease data, exposure-response functions).

For some stressors, the exclusion of health endpoints with insufficient evidence may have led to underestimation of the results, for example for lead and dioxins. On the other hand for dioxins the selection of total cancer as the modelled health endpoint and assuming all cancer cases lethal may lead to overestimation (see also chapters about the individual stressors and the discussion on uncertainty in Chapter 5).

The health endpoints considered in this project and the corresponding exposure-response functions are summarized in Table 3-19 in section 3.12.

Exposure-response functions

For each combination of environmental stressors and health endpoints, exposure-response functions were selected from:

• International recent meta-analyses or WHO guidelines
• If not available: individual high quality studies

Exposure data

Exposure data were as much as possible collected from international harmonized and validated sources. If such data were not available, national data sources were used. In such cases, national data needed to characterize the population exposures in a representative and comparable manner, accounting for potential differences in the urban and rural exposures, different age groups, gender and other relevant sub-groups.

International exposure data were used for SHS, transport noise, ozone, PM and radon. National data were used for benzene, dioxins, formaldehyde and lead, with complementary information from (non-comprehensive) international data sources used when available (AirBase ambient data for benzene; several international multicenter studies for indoor concentrations of benzene and formaldehyde covering some of the participating countries, and WHO Mother’s milk database for dioxins). The sources of exposure data are summarized in Table 3-20 in section 3.12. Exposure data for the target year 2004 are presented in Table 3-21 in the same paragraph. The exposure trends for the year 2010 were estimated using existing data and author judgment to facilitate the evaluation existing policies in the light of the impact estimates for 2004. For several stressors (e.g. lead, dioxins) not enough data were available to make sensible trend estimates. For other stressors (PM, ozone, benzene), temporal and/or spatial variability was so large that reliable evaluations of the trends on the basis of these data were not possible. In these cases, expert judgment was used to estimate trends and corresponding confidence intervals. The estimated trends are summarized in Table 4-4. Due to the large uncertainties no national trend estimates were created.