Multi-symptom asthma is likely to describe a population with more severe disease, as night-time awakenings due to asthma were more common in this group. In addition, the importance of nasal symptoms as risk-factors for multi-symptom asthma is highlighted in this study. Nasal blockage and rhinorrhea, alone and together with allergic rhinitis, were more frequent in subjects with multi-symptom asthma, illustrating that the number of symptoms of rhinitis and severity of asthma are closely associated. Furthermore, symptoms of chronic rhinosinusitis, defined as nasal blockage, rhinorrhea, aching sinuses and/or reduced smell for at least 12 weeks during the last year were closely related to multi-symptom asthma.
When defining multi-symptom asthma, we included individuals reporting physician diagnosis of asthma, use of asthma medication, recurrent wheeze and attacks of shortness of breath, and one more asthma symptom, with the aim of identifying those with more intense disease activity. We suggest that a large component of subjects have a more severe degree of asthma, as they reported much higher frequency of night-time awakenings due to asthma compared with non-asthma and fewer symptom asthma groups. Furthermore, these subjects may represent a group that are "difficult to treat", as they reported several airway symptoms despite having access to asthma medication as required by the multi-symptom asthma definition. Defining severe asthma is not an easy task, as factors such as adherence to treatment, intensity, pathophysiological processes, and the presence of co-morbid conditions, which are clarified in an ATS/ERS statement  and the paper by Redel et al. , must be considered. In the present study, we have decided on using the term "multi-symptom asthma", as it is clearly definable from an epidemiological standpoint. Importantly, no previous attempt has been made to clearly define a group with more severe degree of asthma in previous large-scale population studies, which further illustrates the significance of the present approach. We suggest that our definition of multi-symptom asthma is an appropriate epidemiological tool to define this group of patients with substantially unmet needs .
The prevalence of rhinitis in the general population from the West Sweden Asthma Study, including reported allergic rhinitis, nasal blockage and rhinorrhea, was 37% . However, in both the fewer- and multi-symptom asthma groups, the prevalence of allergic rhinitis increased to approximately 65%, which is in line with previous reports . Thus, the presence of allergic rhinitis was not different between the two groups with different degree of asthma severity, whereas the presence of rhinitis is a clear risk factor for having asthma per se.
Importantly, the prevalence of nasal blockage and rhinorrhea was more than twice as high in the multi-symptom asthma population compared with fewer-symptom asthma, and approximately four times higher in the multi-symptom asthma population versus the non-asthma population. It is especially clear that the prevalence of several rhinitis symptoms was substantially higher in the multi-symptom asthma population, strongly arguing that number of nasal symptoms indeed is closely related to the severity of asthma, even though the prevalence of allergic rhinitis per se does not predict asthma severity. The two strongest risk factors for multi-symptom asthma versus fewer-symptom asthma identified in this study were nasal blockage and rhinorrhea. This is in agreement with clinically recruited cohorts , reporting that severe rhinitis is often associated with more severe asthma. Our study therefore strengthens these previous findings by confirming the close association between severity of rhinitis, and severity of asthma in general, in a random population, and, in addition, clarifying the true prevalence of these symptoms as well as the associations.
As nasal blockage is common in chronic rhinosinusitis, we determined the co-existence of symptoms of this disease with multi-symptom asthma. Indeed, any sign of chronic rhinosinusitis, defined as being present for more than 12 weeks a year, were more frequently reported in the population with multi-symptom asthma compared with both the non-asthma and fewer-symptom asthma groups. Interestingly, more than 60% of subjects with multi-symptom asthma had at least one sign of chronic rhinosinusitis, arguing that a close relationship exists between these conditions. Signs of chronic rhinosinusitis were also associated with multi-symptom asthma regardless of whether the individual reported allergic rhinitis or not, arguing that the allergic status of the individual may be unimportant for this interaction. However, clinical studies that investigate the sensitisation status in patients with signs of chronic rhinosinusitis and multi-symptom asthma are needed to confirm any such independence. An alternative hypothesis could be that infectious agents, including both viruses, bacteria and fungi, could interfere with both nasal symptoms and the severity of asthma .
In addition to the number of nasal symptoms, several other factors appear to distinguish the multi-symptom and fewer symptom asthma populations. A family history of allergy or both allergy and asthma increased the risk of having multi-symptom disease, although a family history of asthma did not clearly distinguish the two categories. In addition, old age, occupational exposure to gas, dust or fumes, and female gender are related to multi-symptom asthma, confirming the involvement of multiple factors for developing a more severe type of asthma. Previous risk-factor analyses of severe asthma have seldom been based on random samples, but rather on clinical cohorts, which lead to substantial selection bias in the analysis .
The strengths of the present study are that it has utilised well-validated epidemiological questionnaires, and it includes a very large random population, which contributes to high internal validity. The response rate was similar or higher than some other international studies of similar nature , albeit slightly lower than some other Swedish studies . Importantly, a survey of those in the current study who did not respond to the questionnaire revealed no differences in prevalence of respiratory symptoms between responders and non-responders, and identical risk-factor profiles . Nevertheless, a relative weakness of any study using postal questionnaires is that that all symptoms and diagnoses are self-reported, which introduces an uncertainty regarding the exact objective clinical diagnosis in each individual. However, the question "have you been told by a doctor that you have asthma" has proven to have very high specificity in Swedish samples . Importantly, the questions used in this study about symptoms of chronic rhinosinusitis have recently been assessed, showing that answers were reasonably stable over time and between countries, were not influence by the presence of allergic rhinitis, and appeared suitable to determine prevalence of chronic rhinosinusitis in epidemiology (unpublished results, submitted for publication). Lastly, understanding and diagnosing chronic rhinosinusitis remains elusive, as epidemiological tools and clinical tools are poorly validated, and the pathophysiological processes are still poorly understood . However, attempts to identify individuals with chronic rhinosinusitis in an epidemiological setting remains a high priority, and further phenotyping of these individuals will require detailed clinical investigations, which is beyond the scope of any large epidemiological approach to identify risk factors.