Different designs can be used to analyze the relationships between respiratory mortality and long term exposure to atmospheric pollution: epidemiological studies (cohort, prevalence study) demonstrate the reality of the relationship and toxicological studies explain it. Cohort studies have the advantage of being able to take into account many confounding factors and thus avoid biases (which is not the case with prevalence studies), but require significant human and financial resources. They were first adopted in the US, but are now more often applied in Europe. The results are relatively consistent, as they all show a statistically significant association between an increase in particulate pollution and cardiopulmonary mortality. Mortality from lung cancer is also associated with long term exposition to particles and sometimes to ozone or nitrogen oxides. Cerebrovascular diseases and sudden death of young children have also been associated with particulate pollution. The relationships are more powerful for long term than short term exposure but are also linear and without threshold. In order to explain these effects (today the causality of the relationship is certain) there are many possible factors, particularly regarding particulate exposures: an increase in cardiovascular risk biomarkers (fibrinogen, white blood cells, and platelets), atherosclerosis, chronic inflammation of lung tissues increased by acute exposure, etc. More and more studies address the interaction between gene and environment and even epigenetic phenomena which could be responsible of these effects. Public Health impact could be quantified. The European E&H surveillance program Apheis, for example, estimated that if PM2.5 levels remained below 15 microg/m(3), a 30 year old person could see his life expectancy increased by 1 month to 2 years, depending on the studied city. Finally, mortality is not the only relevant indicator for health effects of air pollution. ISAAC studies address asthma, allergic rhinitis and eczema among children.