This study evaluated the association of BMI and systemic levels of leptin and adiponectin with levels of exhaled NO and exhaled 8-isoprostanes in subjects with stable asthma and healthy controls. In subjects with asthma, BMI and the systemic leptin/adiponectin ratio were independently associated with a reduction in exhaled NO, and BMI was associated with increased levels of exhaled 8-isoprostanes. In contrast, these associations were not observed in healthy controls with a similar weight distribution. To our knowledge, this is the first study describing the association between BMI, leptin and adiponectin with exhaled NO and exhaled 8-isoprostanes in adults with asthma and in healthy controls.
Although there is compelling epidemiological evidence to support an association between obesity and asthma, plausible mechanisms for this association remain poorly understood. It has been proposed that either obesity-related changes in adipokines and/or the chronic systemic inflammation in obesity, could lead to a parallel increase in airway inflammation. Exhaled NO, a sensitive biomarker of airway inflammation in asthma, would therefore be expected to be higher in obese versus non-obese asthmatics [
6]. However, studies on BMI and exhaled NO do not clearly support this assertion. Some studies have found a positive correlation between BMI and exhaled NO in healthy adults [
30,
31], whereas others have reported no differences between obese and non-obese asthmatic children [
32]. In contrast, our results showed that exhaled NO was inversely associated with BMI, after adjusting for potential confounders in stable asthmatics. Our results are similar to the findings from Barros et al, that showed a negative association between BMI and exhaled NO (
r2 = -0.32 vs.
r2 = -0.35 in our study) in 297 non-smoking asthmatics with a mean BMI of 26 (95% C.I. 25.4 – 26.5) after controlling for potential confounders [
23]. The negative association between BMI and exhaled NO does not necessarily imply that increasing BMI leads to less airway inflammation; it could imply however, that increasing BMI could lead to changes in baseline airway NO redox metabolism, through an increase in baseline airway oxidative stress. In the presence of increased reactive oxygen species, airway NO can be readily converted into reactive nitrogen species (RNS) [
33]. Because the total measured exhaled NO is the end product of NO produced – NO consumed, an increase in the RNS/NO ratio would result in lower measured exhaled NO levels [
34].
Our data support this hypothesis by showing a significant association of BMI with exhaled 8-isoprostanes. Studies have shown that asthmatics have higher levels of exhaled 8-isoprostanes than non-asthmatics, and that levels increase with asthma severity [
35-
37]. In addition, BMI has been associated with higher levels of plasma and urinary levels of 8-isprostanes in men and women [
38,
39]. In contrast to the inverse association between exhaled 8-isoprostanes and exhaled NO observed in our study, Montushi et al reported a positive correlation [
37]; however, this correlation was observed in only 12 mild asthmatics that were not on inhaled corticosteroids, and was not reported for patients with more severe disease that were on inhaled corticosteroids; further, no data on body weight was provided. It is possible that the association between exhaled 8-isoprostanes with exhaled NO changes, depending on whether or not this association is determined during an asthma exacerbation or during baseline conditions.
Alternatively, increasing BMI may lead to an increase in airway oxidative stress via obesity-related changes in adipokines. For example, Leptin increases proportionately with BMI and has been shown to produce reactive oxygen species (ROS) [
40,
41] through multiple mechanisms including, endothelin-1 receptor activation, reduced Nicotine Adenine Dinucleotide (NAD(p)H) oxidase activation, and production of Tumor Necrosis Factor alpha (TNF-α) [
42-
44]. Further, leptin levels in the bronchoalveolar lavage fluid are increased in obese mice models, and instillation of leptin in the airway is associated with acute lung injury in the presence of hyperoxia [
45]. In contrast, adiponectin is inversely associated with biomarkers of inflammation and with BMI [
19,
46], and low levels of adiponectin have been associated with increased systemic oxidative stress, and reduced NO production from endothelial cells [
47,
48].
It is possible that the amount of oxidative stress necessary to shift the airway redox balance towards conversion of airway nitric oxide into RNS exists only when there is a certain balance of leptin and adiponectin. For example, obesity leads to increased leptin and reduced adiponectin; this obesity-induced state of hyperleptinemia and hypoadiponectinemia is associated with increased systemic inflammation and oxidative stress [
46,
49]. It is therefore possible that reaching a certain threshold of obesity-related systemic oxidative stress also results in increased airway oxidative stress. In our study, the ratio of leptin to adiponectin was not associated with exhaled levels of 8-isoprostanes, which may be indicative that other adipokine-independent pathways exist in increasing BMI-related airway oxidation.
This study has some important limiting features. First, the majority of our study population was African American, female, either overweight or obese, and previously diagnosed with moderate to severe disease. These characteristics may limit the external validity of our study, particularly since women appear to be more susceptible to obesity-mediated increase in asthma severity or asthma incidence [
1]; further, these results may not apply to asthmatics with milder forms of asthma severity. Second, determination of causation is impossible and determination of specific mechanisms is difficult in an observational study design. Our results do provide a mechanistic hypothesis by which obesity relates to asthma; however, these results cannot provide information as to why obesity increases the risk for asthma incidence. Third, using questionnaires to evaluate conditions such as obstructive sleep apnea and GERD has a lower sensitivity; therefore, our results could be affected by residual confounding from these misclassified co-morbid conditions. We have to also consider that un-measured confounders, including a more detailed assessment of glycemic control might have important effects on the magnitude of both systemic and airway oxidative stress. Fourth, due to the fact that asthmatics were on several inhaled medications, it is difficult to determine the extent these medicines affected our results. Though all asthmatics were on inhaled corticosteroids, the various dosages and the variable effects they may have on individual patients may confound our results. However, in the study by Barros et al [
23], the association between BMI and exhaled NO was not attenuated when adjusting for inhaled corticosteroids. Further, in the 4-state U.S. National Asthma Survey (NAS) the proportion of asthmatics using ICS is not higher among the obese. In the NAS the use of inhaled corticosteroids among 1059 normal weight asthmatics was 28%, compared to 30% in 985 overweight and 34% in 1015 obese subjects with asthma (p = 0.09) (unpublished observation) [
50]. Fifth, plasma adipokines and exhaled 8-isoprostane levels were only available in 50 and 56 out of 65 patients respectively, and could be a potential source of bias; however, we would expect this bias to be small if any, as the absence of these biological samples was not systematic, and was a consequence of random patient refusal. Sixth, although obese and overweight subjects had higher mean Juniper ACQ, we were not able to detect meaningful differences in asthma severity across BMI categories, given the size of patient population; however, our intention was to minimize differences in asthma severity across BMI categories, to assure that the association between BMI and airway biomarkers were not biased by differences in asthma severity. Seventh, our study is limited by determining airway oxidation stress using 8-isoprostanes using a commercial EIA kit which, although this method is highly specific and has been validated by gas chromatography, there have been contradictory results in the reproducibility of this essay [
51]. Further; we did not explore other airway biomarkers of inflammation and reactive nitrogen species. Finally, we did not assess directly the atopy status in the controls, and are therefore unable to determine how underlying atopy affected the comparison of NO across adults with and without asthma.