BMI but not central obesity predisposes to airway closure during bronchoconstriction

Abstract

Background and objective:

Obesity produces restrictive effects on lung function. We previously reported that obese patients with asthma exhibit a propensity toward small airway closure during methacholine challenge which improved with weight loss. We hypothesized that increased abdominal adiposity, a key contributor to the restrictive effects of obesity on the lung, mediates this response. This study investigates the effect of body mass index (BMI) versus waist circumference (WC) on spirometric lung function, sensitivity to airway narrowing and closure, and airway closure during bronchoconstriction in patients with asthma.

Methods:

Participants underwent spirometry and methacholine challenge. Sensitivity to airway closure and narrowing were assessed from the dose response slopes of the forced vital capacity (FVC), and the ratio of forced expiratory volume in one second (FEV1) to FVC, respectively. Airway closure during bronchoconstriction (closing index) was computed as the percent reduction in FVC divided by the percent reduction in FEV₁ at maximal bronchoconstriction.

Results:

A total of 116 asthmatic patients (56 obese) underwent methacholine challenge. Spirometric lung function was inversely related to WC (p <0.05), rather than BMI. Closing index increased significantly during bronchoconstriction in obese patients and was related to increasing BMI (p =0.01), but not WC. Sensitivity to airway closure and narrowing were not associated with BMI or WC.

Conclusion:

Although WC is associated with restrictive effects on baseline lung function, increased BMI, rather than WC, predisposes to airway closure during bronchoconstriction. These findings suggest that obesity predisposes to airway closure during bronchoconstriction through mechanisms other than simple mass loading.

Summary at a Glance:

The effect of BMI and waist circumference (WC) on spirometric lung function and airway closure during induced bronchoconstriction was assessed in patients with asthma. WC, but not BMI, was associated with restrictive effects on baseline spirometry. However, during bronchoconstriction, airway closure was associated with increased BMI, rather than WC.

INTRODUCTION

Obesity and asthma have both increased in incidence during the past three decades. Although not conventionally thought of as sharing a common pathophysiology, evidence is mounting that they are associated in some way. A meta-analysis of several prospective studies revealed that a high body mass index (BMI) is associated with increased odds of incident asthma.¹ Furthermore, obese individuals with asthma typically present with more severe symptoms than their lean counterparts,² and do not respond as well to controller medication.^3–6 While the epidemiology is striking, the mechanisms linking obesity and asthma are less clear.

One plausible explanation for the relationship between these two conditions is mechanical compression of the respiratory system by excess adipose tissue. Excessive accumulation of fat in the thorax and abdomen substantially alters the mechanics of the lungs and chest wall.^7–10 Consequently, the operating volume of the lungs is reduced;¹¹ this leads to airway narrowing and closure,^12–14 and airway hyperresponsiveness (AHR).¹⁵

Although BMI is a simple and convenient method for classifying the severity of obesity, it does not distinguish between central (abdominal) and peripheral (hips and thighs) adiposity. It also does not account for the pattern of distribution of visceral versus subcutaneous fat. This distinction is important because central obesity—which can be assessed by waist circumference (WC)¹⁶—likely has a direct mechanical effect on pulmonary function. Indeed, central obesity is a marker of visceral adipose tissue,¹⁷ and is a critical determinant of a restrictive (rather than obstructive) spirometric pattern.^{16, 18, 19} On the other hand, general obesity is typically associated with a chronic state of low-grade inflammation,²⁰ and this likely contributes to increased airway wall thickness,²¹ increased baseline airway tone,²² and increased vascular congestion of the airway wall.²³

We previously reported that obese individuals with asthma demonstrate an increased tendency toward small airway closure during bronchial challenge which improved with weight loss, particularly in patients with late-onset disease.^{12, 24} We hypothesized that this was related to the restrictive effects of obesity on the lung. Here, we investigate the effect of BMI versus WC on baseline spirometric lung function, sensitivity to airway closure and narrowing, and airway closure during induced bronchoconstriction in a diverse population of adults with asthma. We hypothesized that if obesity reduced lung function due to mass loading, then increased WC would be associated with increased sensitivity to airway closure and greater degree of airway closure during bronchoconstriction.

METHODS

Study design

The current study involves data from the Study of Asthma and Nasal Steroids (STAN), a multicenter, randomized, placebo-controlled, double-blind trial conducted by the American Lung Association—Airways Clinical Research Centers Network (NCT01118312 at ClinicalTrials.gov).²⁵ Full details of the STAN trial have been described elsewhere.²⁵ Briefly, this 24-week trial investigated whether treatment of chronic sinonasal disease with intranasal corticosteroids improved asthma control in poorly controlled adults and children with asthma. As part of the initial screening for the study, participants underwent an extensive evaluation of the severity of their asthma symptoms and control. Assessments of spirometric lung function and airway hyperresponsiveness to methacholine (AHR) were performed. The study was approved by the institutional review board at each of the 19 centers, and written informed consent obtained from all participants.

Inclusion and exclusion criteria

The current study was limited to adults aged 18 years and older with evidence of AHR. All participants had a physician diagnosis of asthma. Other inclusion criteria included the presence of chronic sinonasal disease defined by a mean score ≥ 1 on the Sinonasal Questionnaire (SNQ),²⁶ and poorly controlled asthma defined by a score of ≤ 19 on the Asthma Control Test (ACT).²⁷ Participants with > 10 pack year smoking history were excluded from the study. Also, patients using systemic or nasal steroids, and patients with other significant co-morbidities were excluded.

Questionnaires

Participants were asked to complete the ACT,²⁷ and the Asthma Symptom Utility Index²⁸ for assessment of their asthma control, and the severity of their asthma symptoms, respectively. Participants were also asked to complete the SNQ to confirm the presence of chronic sinonasal disease.²⁶

Anthropometric measurements

Weight and height were measured with a calibrated scale and stadiometer and BMI calculated. WC was measured as the circumferential distance around the midpoint between the lowest rib and the iliac crest. Hip circumference was measured at the level of the greater trochanters and used to calculate waist-to-hip ratio (WHR). BMI was taken as a global measure of adiposity, while WC was taken as an index of central obesity.^{29, 30}

Assessment of lung function and airway hyperresponsiveness

Spirometric evaluation of lung function and AHR to methacholine were conducted (Koko, Ferraris Respiratory Inc, Louisville, CO) according to the American Thoracic Society guidelines.^{31, 32} AHR was calculated using the dose-response slope (DRS) for forced expiratory volume in one second (FEV₁), as the percent reduction in FEV₁ at the end of the challenge divided by the cumulative methacholine dose in μmol.^{33, 34} A reduction > 4.4% in the FEV₁/μmol ratio (20% decrease in FEV₁ at < 16 mg/mL) was considered a positive response and used as evidence of AHR.¹²

Sensitivity to airway narrowing was computed as the percent reduction in the ratio of FEV₁ to forced vital capacity (FVC) at the end of the challenge divided by the highest dose of methacholine, which is the DRS for FEV₁/FVC ratio. Sensitivity to airway closure was computed using the DRS for FVC, as the percent reduction in FVC at the end of the challenge divided by the highest dose of methacholine. The closing index, a measure of the overall change in FEV₁ attributed to airway closure, was computed as the percent reduction in FVC divided by the percent reduction in FEV₁ at maximal bronchoconstriction.

Statistical analysis

Participants were divided into BMI quartiles as follows: lean (BMI 18.5–24.9kg/m²), overweight (BMI 25–29.9kg/m²), obese (BMI 30 −39.9kg/m²) and severely obese (BMI > 40kg/m²). Participants were also grouped into 4 quartiles according to their WC. Demographic data were summarized with descriptive statistics. Continuous variables were compared using analysis of variance, with log transformation for non-normally distributed variables, while proportions were compared using χ² analysis. Data from the obese group were adjusted for age of asthma onset. We used multiple regression analyses to determine the effects of both BMI and WC on lung function parameters while controlling for race, gender and age. All analyses were performed using STATA 13.0 (StataCorp LLC, College Station TX). P-values < 0.05 were considered statistically significant.

RESULTS

Demographic data and baseline characteristics of study participants

The current study includes 116 adult participants who completed methacholine challenge testing. Many participants in the main clinical trial did not perform methacholine challenge because their FEV₁ was lower than 70% predicted. Thirty-one percent of overweight and 34% of obese participants had an FEV₁ < 70%, whereas only 19% of lean participants had a low FEV₁ (χ² p = 0.04).

Table 1 presents baseline demographics of the study participants by BMI class. There was a significant difference in race distribution, with a lower proportion of Caucasians in the higher BMI categories. Spirometric indices were not significantly different between the BMI groups, but peak flow rates appeared to be lower in those with BMI > 40 kg/m². The provocative concentration of methacholine that induced a 20% fall in FEV₁, as well as asthma symptoms and asthma control, were similar across BMI categories. There were no significant gender differences in BMI, age of asthma onset, asthma control, or asthma severity (not shown).

Table 1:

Baseline Demographics of Study Population by BMI group

BMI (kg/m2)	< 25	25–29.9	30–39.9	> 40	p
n (female)	37 (24)	23 (14)	39 (29)	17 (13)	0.58
Age (years)	37.4±15.5	36.8±11.8	40.1±14.2	39.6±12.4	0.76
Race					0.002
Caucasian	31	12	16	10
Black	3	9	22	5
Other	3	2	1	2
Age of Asthma Onset (years)	15.9±16.2	12.8±11.6	13.9±13.2	18.8±14.0	0.54
Waist Circumference (cm)	84.7±14.4	91.4±15.3	99.8±13.9	109.5±24.0	< 0.001
FEV1 (% pred)	88.2±13.2	91.2±15.0	88.5±17.1	86.9±12.1	0.82
FVC (% pred)	96.8±13.3	103.5±15.8	97.6±15.5	94.9±12.1	0.22
FEV1/FVC (% pred)	91.4±9.2	88.3±8.7	90.5±8.7	91.9±8.7	0.52
Peak flow (% pred)	85.9±15.3	90.4±18.9	89.5±16.8	77.7±18.1	0.08
PC20 (mg/ml)	4.23±5.37	3.50±5.46	4.26±4.79	3.84±6.08	0.44‡
ASUI (range 0–1)↑	0.71±0.20	0.73±0.15	0.73±0.19	0.67±0.19	0.70
ACT (range 5–25)↑	15.7±2.9	16.1±3.1	15.8±3.2	14.3±3.6	0.33

BMI, body mass index; FEV₁, forced expiratory volume in one second; FVC, forced vital capacity; FEV₁/FVC, ratio of FEV₁ to FVC; PC₂₀, the provocative concentration of methacholine that induced a 20% fall in FEV₁; ASUI, asthma symptom utility index; ACT, asthma control test

Values shown are mean ± SD,

^↑higher score indicative of better health

^‡p value shown for Kruskal Wallis test

Baseline demographics and lung function of study participants were also compared across WC quartiles: AHR, asthma symptoms, and asthma control were not significantly different among WC quartiles, whereas baseline FEV₁ and FVC were significantly lower in participants with higher WC (Table 2). Similar results were observed with WHR (Supplementary Table S1).

Table 2:

Baseline Demographics of Study Population by waist circumference quartile

Waist Circumference Quartile	1	2	3	4	p
Waist Circumference (cm)	74.9±7.0	88.1±2.5	98.7±3.8	119.6±13.5
n (female)†	30 (27)	27 (14)	28 (20)	26 (17)	0.02
Age (years)	35.3±15.5	35.9±12.1	42.4±13.8	42.7±12.5	0.07
Race					0.60
Caucasian	20	18	16	12
Black	8	7	10	13
Other	2	2	2	1
Age Asthma Onset	15.8±15.7	12.8±12.8	16.0±14.4	15.9±13.8	0.81
BMI	25.0±4.3	30.6±7.5	30.6±7.0	36.3±9.6	< 0.001
FEV1 (% pred)	94.2±13.5	85.8±12.2	91.6±14.1	83.8±16.0	0.02
FVC (% pred)	104.3±14.4	97.6±9.2	99.7±15.3	91.7±15.5	0.01
FEV1/FVC (% pred)	90.8±9.8	87.8±8.8	92.2±7.2	91.3±8.1	0.26
Peak flow (% pred)	85.5±15.0	85.2±19.6	94.3±11.2	85.5±20.1	0.08‡
PC20 (mg/ml)	3.05±4.17	3.67±5.21	6.86±7.15	2.83±3.05	0.27‡
ASUI (range 0–1)↑	0.72±0.19	0.67±0.16	0.72±0.17	0.73±0.20	0.62
ACT (range 5–25)↑	16.2±2.9	14.6±3.4	16.0±3.0	15.4±3.2	0.24

BMI, body mass index; FEV₁, forced expiratory volume in one second; FVC, forced vital capacity; FEV₁/FVC, ratio of FEV₁ to FVC; PC₂₀, the provocative concentration of methacholine that induced a 20% fall in FEV₁; ASUI, asthma symptom utility index; ACT, asthma control test

Values shown are mean ± SD,

^↑higher score indicative of better health

^†111 participants had measurements of waist circumference

^‡p value shown for Kruskal Wallis test

Effect of BMI on spirometric lung function, sensitivity to airway narrowing and closure, and closing index

To determine the effect of global obesity on spirometric lung function, sensitivity to airway narrowing and closure, and the closing index, we assessed the relationship of these parameters to BMI category. BMI was not significantly related to spirometric lung function (Table 1). Sensitivity to airway narrowing and closure were not significantly associated with BMI (Figures 1A and 1B). However, the closing index was significantly different across BMI categories, with significantly higher indices recorded in patients with higher BMI (Figure 1C).

Figure 1:

Relationship between sensitivity to airway narrowing and closure, and closing index with BMI group: 1A relationship between sensitivity to airway narrowing and BMI category, p =0.72, by trend test; 1B relationship between sensitivity to airway closure and BMI category, p =0.94, by trend test; 1C relationship between closing index and BMI category, p < 0.01 by trend test.

Effect of WC on spirometric lung function, sensitivity to airway narrowing and closure, and closing index

To determine the effect of central obesity on spirometric lung function, sensitivity to airway narrowing and closure, and the closing index, we assessed the relationship of these parameters to WC. WC was inversely related to baseline FEV₁ and FVC (Table 2) but not to sensitivity to airway narrowing or airway closure (Figures 2A and 2B), or the closing index (Figure 2C).

Figure 2:

Relationship between sensitivity to airway narrowing and closure, and closing index with waist circumference quartile: 2A relationship between sensitivity to airway narrowing and WC, p =0.95, by trend test; 2B relationship between sensitivity to airway closure and waist circumference quartile, p =0.17, by trend test; 2C relationship between closing index and waist circumference quartile, p =0.85 by trend test.

Multi-variate analysis of combined effects of WC and BMI on spirometric lung function, sensitivity to airway narrowing and closure, and closing index

We investigated the combined effects of BMI and WC on spirometric lung function. When BMI and WC were included in the same model, BMI was associated with increased FVC, whereas WC was associated with decreased FEV₁ and FVC (Table 3).

Table 3:

Multi-Variate Regression Analysis of relationship between anthropomorphic measures and lung function

	Waist circumference*		BMI*
	Coef (SE)	p	coef (SE)	p	R2
FEV1 % predicted	−22.9 (8.8)	0.01	9.41 (6.40)	0.15	0.04
FVC % predicted	−32.4 (8.5)	<0.001	13.7 (6.2)	0.03	0.11
FEV1/FVC % pred	5.49 (5.41)	0.31	−3.10 (3.94)	0.43	0.00
DRS FEV1/FVC*	−0.09 (0.61)	0.89	−0.38 (0.45)	0.40	0.00
DRS FVC*	−0.56 (0.52)	0.29	0.05 (0.90)	0.72	0.00
Closing Index*	−0.31 (0.30)	0.30	0.56 (0.22)	0.01	0.06

BMI, body mass index; FEV₁, forced expiratory volume in one second; FVC, forced vital capacity; FEV₁/FVC, ratio of FEV₁ to FVC; DRS, dose response slope

^*log transformed for regression analysis. WC and BMI were not subject to collinearity.

Sensitivity to airway narrowing and closure were not related to either BMI or WC in univariate analysis, or multi-variate analysis (Table 3). In multi-variate analysis, the closing index was significantly related to BMI but was not related to either WC (Table 3) or WHR (Table S2).

DISCUSSION

We investigated the effect of general obesity (BMI) and central adiposity (WC) on baseline spirometry, sensitivity to airway narrowing and closure, and airway closure during induced bronchoconstriction in individuals with asthma. The purpose of this study was to distinguish the restrictive effects of obesity on lung function from other factors related to obesity. Our principal findings were that (i) WC was the main determinant of decreased spirometric lung function, (ii) sensitivity to airway narrowing and closure were unaffected by BMI or WC, (iii) increased BMI predisposed patients with asthma to airway closure, independent of WC.

The relationship between obesity and asthma has been well documented,¹ but there is a paucity of data on the physiological effects of BMI versus WC on some distinguishing features of asthma. There is a growing appreciation that the term obesity includes many pathophysiological entities which variably affect baseline lung function and asthma.³⁵ Indeed, it is generally thought that obesity simply leads to restrictive effects on lung physiology, but this relationship may not be so straightforward, as illustrated by the current study.

In terms of the effects of obesity on spirometric lung function, we found that baseline FEV₁ and FVC were inversely related to WC but not BMI (Tables 1–2), suggesting that these parameters were more directly related to the restrictive aspects of obesity. This finding is consistent with a large cross-sectional study that found WC predicted FEV₁ and FVC independent of BMI.¹⁸ Unexpected in the current study was that BMI was associated with increased FVC (Table 3). Although BMI and FVC initially appears to be unrelated (Table 1), FVC becomes positively related to BMI in the study population after controlling for both WC and BMI (Table 3). We do not know the reason for this, but studies in children provide some insights—obesity is associated with accelerated lung growth, and FVC is increased out of proportion with FEV₁, producing airway dysanapsis.³⁶ In longitudinal and cross-sectional studies of young adults, increased body weight is associated with moderate increases in FVC.³⁷ This is attributed to the “muscularity” effect and is dependent on age and height. We do not have longitudinal data, but it is possible that differential effects of BMI on lung growth, and WC on lung restriction may both be involved in the effects of obesity on spirometric lung function in adults.

Prior studies have suggested that peripheral airway dysfunction in response to methacholine occurs in obese non-asthmatics,^{22, 38} and we have reported that weight loss improves sensitivity to airway closure in late onset non-allergic asthmatics.³⁹ However, to the best of our knowledge, the present study is the first to investigate whether increased adiposity or body-fat distribution increases sensitivity to airway closure during methacholine challenge. Neither WC nor BMI have a significant effect on sensitivity to narrowing or closure, consistent with the fact that not all studies have found a relationship between BMI and AHR.^{40, 41} In fact, some studies have reported a relationship between BMI and AHR in non-asthmatics but not in asthmatics.^{14, 42}

BMI but not WC contributed significantly to the closing index, a measure of the degree of airway closure during bronchoconstriction. This is in agreement with a recent study by Burgess and colleagues that reported that airway closure – measured by percent fall in FVC rather than the closing index – explained nearly half the increased airway reactivity associated with increased BMI, and this relationship was not altered by including WC in the model.¹⁴ These data suggest that the tendency towards airway closure during bronchoconstriction is not simply a function of the restrictive effects of obesity, but rather is related to factors that track more closely with BMI than WC. Perhaps, this could be related to lung growth, as discussed above. Alternatively, it may relate to factors more closely associated with subcutaneous versus visceral adipose tissue. Circulating leptin is one factor more strongly related to subcutaneous than visceral adipose tissue.^{43, 44} Leptin has several functions that could impact AHR such as involvement in surfactant production, lung development and bronchial tone.^{40, 45, 46}

An important limitation of this study is that it is only comprised of patients with poorly controlled asthma and chronic sinonasal disease who demonstrated FEV₁ > 70%. This resulted in the exclusion of a substantial proportion of overweight and obese participants. It is therefore unclear whether the findings of this study are generalizable to patients with more severe airflow limitation who do not present with upper airway disease. However, since a greater proportion of obese study participants had baseline FEV₁ lower than 70%, we speculate that the effect of increased BMI on airway narrowing and airway closure would be heightened in these subjects.

In summary, we have shown that WC is associated with restrictive effects on baseline spirometry. However, increased BMI, rather than WC, predisposes to airway closure during bronchoconstriction. Taken together, these findings suggest that obesity predisposes to airway closure during bronchoconstriction through mechanisms other than simple restriction. This could be clinically important because increased airway closure during bronchoconstriction in obese subjects could limit deposition of inhaled medications and reduce the efficacy of standard therapeutic approaches.

Abbreviations:

ACT

asthma control test

AHR

airway hyperresponsiveness

ASUI

asthma symptom utility index

BMI

body mass index

DRS

dose-response slope

FEV₁

forced expiratory volume in one second

FVC

forced vital capacity

FEV₁/FVC

ratio of FEV₁ to FVC

PC₂₀

the provocative concentration of methacholine that induced a 20% fall in FEV₁

SNQ

Sinonasal Questionnaire

STAN

Study of Asthma and Nasal Steroids

WC

waist circumference