Prevalence studies, i.e. cross-sectional studies, may provide data about risk factors for diseases however, a found association could be a cause, a consequence of disease or a parallel phenomenon. Thus, longitudinal studies of cohorts of the general population are needed as these studies provide more valid results regarding risks, and the probability of the associations being risks is greater than that of the associations that are derived from cross- sectional studies. These important methodological aspects have been explored in a chapter of this Monograph. One important area that requires consideration is rare respiratory disease. In Europe, a disease or disorder is defined as rare when it affects less than one in 2000 people however, as many as 30 million people may be affected by one of over 6000 rare diseases. It has been shown that many major diseases can be subdivided into individual diseases called phenotypes, some of which are classified as rare. This is often the case with respiratory disease. Rare diseases constitute appropriate models for the underlying mechanisms of other diseases. The specific features associated with rare diseases (low individual patient numbers, diversity, geographical location) mean that research needs to be developed at a large level in order to optimise funding, infrastructures and technological platforms. Patient registries and databases are key to the development of epidemiological and/or clinical research in this area, as they pool data and achieve a sufficient sample size, leading to improvements in patient care and healthcare planning. The creation of a rare respiratory disease registry would also be a powerful tool in creating a network of experts. Due to the increasing difficulty and cost of carrying out new general population surveys, in which a progressively decreasing proportion of people agree to participate, it is likely that there will be rise in the epidemiological use of routinely collected statistics, as well as medical records stored by family practitioners. Data mining procedures within the huge database created by the widespread use of e-health instruments (patient diaries, electronic health cards, etc.) will help. Another important issue that needs to be developed in the future is exposomics. Genetics have been found to account for only around 10% of diseases, and the remaining causes appear to be environmental. In order to understand the causes of respiratory disease and its prevention, the impact of the environment needs to be further investigated. Exposomics can be defined as the measure of all the exposures of an individual in a lifetime and how those exposures relate to his/her health. In respiratory disease, an individual’s exposure begins before birth with passive smoking and includes insults from the environment, like air pollution, diet, lifestyle and occupational sources, and their interactions with characteristics such as genetics, epigenetics and physiology. This requires appropriate methodologies, such as epidemiology, biostatistics, omics and data mining. Respiratory epidemiology also plays an important role in translational research, and molecular epidemiology is increasingly important. Modern translational research covers epidemiology, clinical aspects (including phenotyping or endotyping) and research on cellular mechanism (including inflammation, proteomics and genetics). The contribution of epidemiology can guarantee a representative sample for study. Studies of the gene– environment interaction play an increasing role in modern epidemiological research. We believe that in this issue of the Monograph the reader will find methodological information and content that will be useful in forthcoming calls for research proposals issued by the EC within the framework of Horizon 2020, which includes relevant themes for public health and respiratory medicine. xvi