Obesity induces numerous changes in lung mechanics and function  (Fig. 1). The most obvious change is a reduction of the expiratory reserve volume (ERV), with an associated reduction in functional residual capacity (FRC), from changes in elastic properties of the chest wall as obese subjects breathe at lower lung volumes. Tantisira et al. found an inverse association between BMI and FEV1/FVC ratio, but obesity is not considered to induce any significant airway obstruction [41, 53, 54]. However, when breathing at low FRC, airway calibre is reduced and airway resistance can increase [54-57]. Breathing at low lung volumes may unload airway smooth muscle (ASM) and reduced breathing-related airway distension in the obese can theoretically increase airway responsiveness [41, 58-61]. This is supported by our previous observation that obese non-asthmatic subjects have lost the protective effect of a deep inspiration, observed in non-obese non-asthmatic subjects . However, other mechanisms related to small airway closure may prevail to explain altered airway functions [59, 63]. Although these changes may favour the development of AHR, and increases in the prevalence of AHR in the obese have not been universally reported [33, 42, 64, 65]. Furthermore, bronchoconstriction increases lung hyperinflation in the obese compared with non-obese individuals , these effects being possibly due to increased expiratory flow limitation [66, 67] and increased airway closure .
Severe obesity may be associated to a reduced total lung capacity (TLC), with a restrictive pattern, but the residual volume (RV) usually remains unchanged [47, 68, 69]. Finally, obese subjects show an altered pattern of breathing, with higher breathing frequencies and reduced tidal volumes, even when exercising . Obesity is associated with a reduction in respiratory system compliance, increased work of breathing and increased exercise-induced dyspnoea [53, 69].
Adipocytes may store and release a variety of inflammatory agents and obesity is generally considered as a low-grade systemic inflammatory condition [70, 71]. Increased levels of various inflammatory cytokines and mediators such as tumour necrosis factor-alpha (TNF-α), interleukin-6 (IL-6), eotaxin, vascular endothelial growth factor (VEGF), monocyte chemotactic protein (MCP) or markers of systemic inflammation such as C-reactive protein and serum fibrinogen have been observed in obese subjects . Furthermore, levels of markers of oxidative stress such as 8-isoprostane are increased in the blood and the lungs of obese compared to non-obese subjects .
Another characteristic of obese subjects is the presence of an increased number of leucocytes, particularly cytotoxic T cells and macrophages in the adipose tissue, possibly under the influence of hypoxia, chemoattractants or excess in lipid substrates [74-78]. Adipose tissue resident macrophages can produce a large variety of inflammatory molecules, which are increased in the presence of obesity .
A possible role of energy-regulating hormones has been suggested to explain the influence of obesity on asthma. In this regard, plasma levels of leptin, an IL-6 like adipokine secreted from adipocytes, are increased in obese subjects . Leptin has multiple pro-inflammatory effects, and acts on the innate immune system [80-85], promoting chemotaxis and cytokine and surface markers of activation production [86-93]. Leptin can also increase alveolar macrophage activation [94, 95]. A link has been observed between leptin and asthma, particularly in children . However, although leptin is closely related to obesity, its role in the development of asthma is unclear [97, 98].
Otherwise, the anti-inflammatory hormone adiponectin levels are lower than in non-obese subjects [96, 99]. Its levels increase with weight loss [100, 101]. Adiponectin, an insulin-sensitizing hormone, inhibits the production of pro-inflammatory cytokines by macrophages and monocytes and increases IL-10 and IL-1 receptor antagonist expression [102, 103].
Serum leptin and adiponectin concentrations have been independently associated with asthma in asthmatic subgroups [101-106], but the influence of obesity on asthma does not seem to be mainly mediated through the influence of these adipokines [99, 100]. A study looking at methacholine response in children showed no significant association between serum leptin or adiponectin and asthma . The link between adipokines and asthma seems to be influenced by age and gender and associations are found in prepubescent boys, prepubescent girls, and premenopausal women for leptin and prepubescent girls and premenopausal women for adiponectin [108-111]. Furthermore, Engbers et al. found no correlation between plasma leptin concentration and airway inflammation, measured after an Adenosine Monophosphate (AMP) challenge . In a Scandinavian prospective cohort study of children and adults, incident asthma was associated with obesity and prior weight gain, but there was no association between leptin, adiponectin, C-reactive protein, or insulin with asthma in a multivariate analysis . Finally, Sutherland et al. showed that in obese and normal weight women with and without asthma, obesity was associated with increased interleukin (IL)-4 and IL-6, C-reactive protein, and leptin, and asthma was associated with C-reactive protein and Th2-type cytokines, but there were no obesity-by-asthma interactions . Therefore, there are uncertainties on the link between obesity-related adipokines and airway inflammation or asthma .
In regard to airway inflammation, De Winter et al. showed that a higher BMI is associated with a higher eNO in healthy patients while Kazaks found no significant relationship between exhaled NO and BMI in adult asthma although there was no correction for use of corticosteroids [116, 117]. In another study, BMI and the ratio of leptin to adiponectin were, respectively, associated with reduced levels of exhaled NO .
We (L. P. Boulet, H. Turcotte, unpublished data) and Todd et al. found no significant difference in induced sputum cell differential in obese non-asthmatic compared to non-obese subjects globally  although we and Van Veen et al. found an inverse relationship between BMI and sputum eosinophils [30, 40]. Similarly, Barros et al. observed a negative association between BMI and Fractional exhaled Nitric Oxide (FeNO) in overweight or obese patients with asthma . A recent report from Scott et al. suggested that there is an increase in neutrophilic airway inflammation in obese asthmatic female patients  Furthermore, in cluster analyses, asthma in the obese patient was associated with a non-eosinophilic inflammatory phenotype .
The possible contribution of an increased oxidative stress in obese subjects on airways has also been studied. BMI is associated with increased levels of exhaled 8-isoprostanes  although systemic oxidant stress is not independently associated with asthma, suggesting that it does not explain the obesity-asthma association .
Finally, there are few studies on airway structure in obese patients with asthma. However, airway remodelling features, including smooth muscle mass area, subepithelial fibrosis and epithelium integrity seem to be similar between obese and non-obese asthmatic patients . The existence of fat loculation on bronchial biopsies of morbidly obese subjects was suggestive of airway fat infiltration [63, 123]. The possibility that local fat can be a trigger for a local inflammatory process has, however, to be studied.
Role of comorbidities
Comorbid conditions such as gastro-oesophageal reflux, OSA, diabetes, dyslipidaemia, and cardio-vascular diseases are more prevalent in the obese and may possibly influence the clinical expression of asthma, although a possible role in the development of the disease is not as obvious [28, 46]. In a study by Dixon et al. on 402 subjects (51% obese), those with a higher BMI reported a higher prevalence of reflux symptoms, but the prevalence of pH probe acid reflux was similar in all groups . Reflux was not associated with measures of asthma control in obese patients. Furthermore, symptoms and self-report of OSA were more common with increasing BMI and associated with worse asthma control. So, Gastroesophageal Reflux Disorder (GERD) probably has little influence on asthma control in obese patients while OSA could, although in some studies, adjustment for these two conditions did not change the relationship between asthma and obesity [132, 133].
A possible role of the metabolic syndrome, characterized by the association of type 2 diabetes, systemic hypertension, dyslipidaemia, and central obesity has been suggested to explain the link between asthma and obesity. Leone et al.  found an independent relationship between reduced lung function and metabolic syndrome, mainly due to abdominal obesity and Lee et al. found that the metabolic syndrome was associated with asthma symptoms . Furthermore, insulin resistance, often found in the obese, has been linked to allergen sensitization and asthma symptoms [136, 137].
Patients with morbid obesity (BMI > 40 kg/m2) have an increased prevalence of comorbid conditions such as hypertension, diabetes, congestive heart failure, asthma, poor functional status, sleep apnoea, pulmonary hypertension, venous thromboembolism, or venous oedema compared to leaner patients . However, how comorbid conditions contribute to the development or changes in the clinical expression of asthma in the obese subject remain to be explored.