The primary goal of asthma treatment, according to the Spanish guidelines for asthma management (GEMA 2009), is to achieve and maintain disease control,1 which implies normal health-related quality of life.2 However, in a high number of patients, this objective is not usually achieved. A study by González-Barcala et al3 conducted in the field of primary care on a sample of 2159 asthmatics showed that 64% were poorly controlled.

There are many factors that explain poor control of asthma: sub-optimal doses of medication,4 poor compliance or inadequate inhalatory tecnique5 and an asthmatic phenotype that do not usually respond to medication6 are examples. In two series of patients with difficult to control asthma, concomitant processes were found that could aggravate the disease in 19–34% of cases.7,8 It is possible that the presence of comorbidity is associated with a poorer therapeutic response to the escalated treatment recommended for patients with poorly controlled asthma, but it is clear that the specific treatment of concurrent processes improves asthma control.8

In recent years, using a standardised methodology, instruments have been developed that are designed to measure asthma control from the perspective of the patient. Among them, the Asthma Control Test (ACT),9 whose Spanish-language version has been recently validated,10 has emerged as a simple and useful questionnaire for assessing the clinical status of asthmatics. Meanwhile, asthma specialists have been interested in finding an easily measurable and reproducible objective to facilitate monitoring of the disease and to avoid over or under treatment. Exhaled nitric oxide (FENO), a surrogate indicator of eosinophilic bronchial inflammation, has been used to guide asthma therapy.11,12 The relationship between the two types of instruments, clinical and biochemical, has rarely been studied and it has been suggested that it is possible that they measure different aspects of the disease.13

This study is based on the hypothesis that asthma control may be adversely affected by concurrent processes. Our primary objective was to know whether poorly controlled asthma is related to the presence of comorbidities. As a secondary objective, we studied whether the degree of control, measured by the ACT, correlates with FENO levels.

Measurement

The results of this study show the high percentage (>90%) of asthmatic patients who have comorbidities likely to worsen asthma control. Consequently, there are no significant differences in incidence between controlled asthmatic patients and those who are not (95% vs 97%) if comorbidity is defined as the presence of at least one potentially aggravating factor. However, the study shows that the coexistence in the same patient of several comorbidities is more common in cases of poorly controlled asthma (23.2% of controlled patients showed three or more comorbidities, while the same occurs in 44% of uncontrolled patients.)

There is increasing interest in the influence of comorbidity on the development and control of asthma17 and it has been suggested that the identification of these factors be part of the routine evaluation of asthma patients, particularly those that have difficulty controlling their asthma. An association between asthma severity and response to treatment and the following conditions has been reported: gastro-oesophageal reflux,18,19 obesity,20,21 vocal chord dysfunction,22 anxiety-depression,23,24 rhinosinusitis25,26 and smoking habit.27 However, the influence of these diseases on the course of asthma is variable and, in many cases, is not clear at all. Theoretically, it could alter the asthma phenotype (e.g. tobacco smoking, ABPA), be part of the same pathophysiological process (e.g. nasosinusitis) or could act as confounding factors in the diagnosis or assessment of asthma control (e.g. anxiety and depression). However, the impact of any of these processes in each individual patient is different and therefore difficult to predict. Our results show that the vast majority of asthmatics have some concomitant disease that can affect asthma evolution, whatever may be the degree of control. Only nasal polyposis (8.8 vs 0%), gastro-oesophageal reflux (22.8 vs 8.9%) and ABPA (9.8 vs 0%) were more frequent in uncontrolled patients (depression did not reach statistical significance by a small margin.) Patients with poor control show a clear trend towards having more comorbidities at the same time.

From these results we can get two obvious practical consequences. On the one hand, if the presence of comorbidities is associated with poor asthma control, the specific treatment of them could result in a benefit in patient quality of life. However, this is not always the case, at least not in the case of subclinical gastro-oesophageal reflux diagnosed by pH monitoring. The study of The American Lung Association Asthma Clinical Research Centers showed that treatment with esomeprazole for asthmatics with this concomitant process did not improve control of the disease.28 On the other hand, the co-existence of processes that potentially exacerbate asthma, but that have a different biological nature, can make an increase in medication based on the recommendations of the GEMA ineffective because the patient's symptoms would not be directly related to the bronchial inflammatory process. If this is the situation, we find cases of excessive and inappropriate treatment. At this point it would be very useful to have a marker to predict response to treatment, distinguishing between under-treated patients and those for whom increasing specific therapeutic strategies for asthma would be superfluous. Our group has shown that elevated FENO values imply that asthma control can be achieved with an increase in medication. In contrast, a poorly controlled patient with normal values of FENO is unlikely to improve with further treatment for asthma,29 which would result in a decrease in therapeutic expectations, at least if we base the concept of asthma control on symptom questionnaires.

The ACT has proved to be a reliable instrument to measure asthma control.9 However, it is based on the subjective perception that a patient may have regarding their symptoms and the answer to any of the questions included. For example, “during the past four weeks, how often have you noticed that you are short of breath?” can be influenced by concurrent processes that can cause similar symptoms (e.g., obesity, anxiety, etc.) Thus, it is not surprising that we found no correlation between ACT values and the FENO values. The same conclusion was reached by Shirai et al in a study with fewer patients and which did not emphasise the presence of comorbidity.13 Since they found high levels of FENO in patients with asthma control, they speculated that the presence of underlying eosinophilic inflammation (what FENO would indirectly show) is not an obstacle for the medication to eliminate symptoms. In our series it was observed that the FENO value was significantly higher in the group of patients with sub-optimal control (43.1 vs 24.5ppb), and the finding of cases with high FENO and an ACT≥20 was very rare. Although the hypotheses of Shirai were true, we believe it is too simple to explain the discrepancy between FENO and ACT. Patients with difficult to control asthma are a heterogeneous group and it is possible that our series included neutrophilic phenotype individuals or patients whose symptoms corresponded to the presence of comorbidity, both situations in which the ACT could be low and the FENO normal.

The originality of this study is that we have arranged the groups according to the degree of asthma control, something that had not been done in previous studies which focused on the potential effect of comorbidity on asthma. This allowed us to bring our approach to the objectives of the latest version of the GEMA – targeting control as a desirable objective – and to speculate on the potential limitations of the ACT, a tool widely used in clinical practice. Establishing a causal link between the presence of comorbidity and poor control of asthma will not be easy, as there are potential misleading factors that make the results difficult to interpret.

A potential limitation of this study is the number of patients that, although higher than in similar studies, may not be sufficient to demonstrate a statistical relationship between asthma control and certain processes that, like depression, show a tendency to reach significance in our results. Another shortcoming lies in the fact that the diagnosis of gastro-oesophageal reflux was purely clinical and pH monitoring or endoscopy was performed in all cases, something that presumably would have allowed the identification of sub-clinical cases.

The findings of this study are: the vast majority of patients with asthma, be it well or poorly controlled, present a process that is likely to worsen the control of the disease; patients with sub-optimal control show a tendency to accumulate comorbidities; there is no correlation between ACT and FENO, meaning that these tools reflect different clinical realities.