In this issue of Monaldi Archives, Al-Muhairi et al. describe the epidemiology of Communityand Hospital-acquired pneumonia in United Arab Emirates over five years [1].Most of the cases (287, 80%) were actually community-acquired (CAP) and only 74 hospital acquired (HAP). Considering the five-year observation only a few HAP cases each year were observed and no clear conclusion can be drawn from this series.More interesting are the observations that the annual admission rate for CAP increased from 250/100,000 admissions in 1997 to 710/100,000 in 2002 and the presence of a seasonal trend in the admission rate.The lack of any established rules for hospital admission can only partially explain the increase in the number of patients hospitalised. A change in population risk factors or in CAP etiology can also be involved but, unfortunately, these variables are not addressed by this study.For example, in Al-Muhairi study, as commonly reported in the majority of studies into pneumonia, there is a large proportion of cases that have no pathogen identified, mainly because the appropriate tests were not performed [2]. In this study less than 30% of patients had sputum examination, less than 50% blood culture and only 14/287 had serology tests. This is a major pitfall of the study, that cannot give any indication on CAP etiology in the examined aerea. The seasonal differences in CAP incidence seem to be clearly related to population migration. This is an interesting point, rarely addressed in literature. The authors suggest that the peak of pneumonia cases follows the return from Asia of the expatriate population, this means a transmission of infections from Asian residents to local population. What is not clear is that only 3% of admitted patients are Asians. This may imply a high rate of infection spreading between Asian immigrants and the local population, underlying the importance of etiologic studies in this setting. Both increasing incidence and seasonal variation should foster the authors to design a new study to better investigate the epidemiological risk factors and the etiology of CAP in their population.Another interesting article in this issue of Monaldi Archives addresses the effect of cigarette smoking on hydrogen peroxide (H 2 O 2 ) and thiobarbituric reactive substances (TBARs) concentrations in exhaled breath condensate (EBC) in patients with CAP [3].The effects of tobacco smoke on the airway mucosa are well known. The chronic inhalation of smoke favours the adhesion of S. pneumoniae and H. influenzae to the buccal epithelium. Smoking alters mucociliary transport, humoral and cellular defences, and epithelial cells. The proportion of smokers in hospitalised pneumonias is generally increased, which would indicate a role of smoking in the acquisition of infection.In the Stolarek et al. study the different responses to oxidative stress in smokers and nonsmokers with CAP is addressed. The study relies on the use of EBC, an interesting non-invasive technique for obtaining samples from the lungs. In EBC large num...