ARTICLE IN PRESS

Respiratory Medicine (2006) 100, 648–657

KEYWORDBody massObesity;Asthma coAsthma qulife;Asthma seAsthma ma

0954-6111/$ - sdoi:10.1016/j.r

$SupportedFondation de l’�Correspond

West, MontrealE-mail addr

Higher BMI is associated with worse asthma controland quality of life but not asthma severity$

Kim L. Lavoiea,b,�, Simon L. Bacona,c, Manon Labrecquea,Andre Cartiera, Blaine Dittod

aResearch Center, Department of Chest Medicine, J-3190, Hopital du Sacre-Cœur de Montreal,5400 Gouin West, Montreal, Quebec, Canada H4J 1C5bDepartment of Psychology, University of Quebec at Montreal (UQAM), P.O. Box 8888,Succursale Center-Ville, Montreal, Quebec, Canada H3C 3P8cDepartment of Exercise Science, Concordia University, 7141 Sherbrooke St West, Montreal, Quebec,Canada H4B 1R6dDepartment of Psychology, McGill University, 1205 Dr. Penfield Ave., Montreal, Quebec, Canada H3A 1R1

Received 20 May 2005; accepted 2 August 2005

Sindex;

ntrol;ality of

verity;nagement

ee front matter & 2005med.2005.08.001

by a fellowship grant frHopital du Sacre-Coeuring author. Research Ce, Quebec, Canada H4Jess: kim-lavoie@crhsc.r

Summary Asthma and obesity tend to co-occur, but relatively few studies havelinked obesity, measured using body mass index (BMI), to clinically relevant measuresof asthma morbidity. This study assessed BMI in a Canadian sample of asthmaoutpatients, and evaluated associations between BMI and levels of asthma severity,asthma control, and asthma-related quality of life.

A total of 382 adult asthma patients underwent demographic and medical historyinterviews on the day of their clinic visit. Patients’ self-reported height and weightwere used to calculate BMI (kg/m2). Asthma severity was classified according theGINA (2002) guidelines. Patients completed the Asthma Control (ACQ) and AsthmaQuality of Life (AQLQ) Questionnaires and underwent standard pulmonary testing(spirometry).

A total of 139 (36%) patients had a normal BMI; 149 (39%) patients wereoverweight; and 94 (25%) patients were obese. There was no relationship betweenBMI and asthma severity when controlling for age and sex. Patients with higher BMIscores had higher ACQ and lower AQLQ scores, independent of age, sex and asthmaseverity.

Elsevier Ltd. All rights reserved.

om the Canadian Institutes of Health Research (KLL) and a Young Investigators grant from thede Montreal (Auger Foundation).nter, Department of Chest Medicine, J-3190, Hopital du Sacre-Cœur de Montreal, 5400 Gouin1C5. Tel: +1 514 338 2222 3709; fax: +1 514 338 3123.tss.qc.ca (K.L. Lavoie).

ARTICLE IN PRESS

BMI, asthma severity, asthma control and quality of life 649

Results identify higher BMI and obesity as potential behavioral factors related toworse asthma control and quality of life, but not asthma severity, and suggestimportant avenues for asthma management and control initiatives.& 2005 Elsevier Ltd. All rights reserved.

Introduction

Despite important advances in diagnosis and treat-ment, the prevalence of asthma has increased inrecent years in both Canada and the United States.Asthma currently affects over 8.4% of Canadians1

and 7% of Americans.2 Between 1980 and 1994, thenumber of people with asthma increased by over100%.3 By 1999, over 2.17 million Canadians1 andnearly 15 million Americans4 had asthma. In 1990,medical expenditures for asthma-related healthcare costs exceeded $6.2 billion in the US5 and $648million in Canada.6 In 2001, Canadian statisticsshowed that only 43% of asthmatics were wellcontrolled according to Canadian Consensus Guide-lines.7 This indicates that more than half of allasthmatics in Canada have poorly controlled oruncontrolled asthma and are experiencing symp-toms at a level that interferes with daily activitiesand reduces overall quality of life.

In the last 3 decades, the prevalence of obesityhas also risen in both the US and Canada. Recent(1998) statistics estimate the rate of obesity,defined as having a body mass index (BMI) of30 kg/m2 or greater, at over 30% in the US8–10 and15% in Canada.11 When you also consider those whomay be considered overweight (BMIX25 kg/m2),these rates as much as triple, reaching 65% ofAmericans8 and nearly 50% of Canadians.11 Inabsolute numbers, over 11 million Canadians and130 million Americans are either overweight orobese, a phenomenon that costs the US healthcaresystem an estimated $70 billion dollars per year12,13

and the Canadian healthcare system as much as$1.8 billion per year.14 The health consequences ofobesity are varied and well established. Obesity hasbeen shown to either increase risk for or exacer-bate many serious illnesses including hypertension,diabetes, arthritis, hypercholesterolemia, coronaryartery disease, restrictive lung disease, sleepapnea, and cancer.10,13,15

The concurrent rise in asthma and obesity rateshas prompted researchers to investigate associa-tions between the two disorders. Cross-sectionalstudies have found evidence of an associationbetween asthma and obesity in both children16–18

and adults,19–21 though some have suggested amore consistent link between asthma and obesity inwomen.20,21 Few studies have investigated the

temporal relationship between asthma and obesity,however, a recent prospective study of 85,911women found that obesity and increasing BMI wasrelated to higher rates of new adult-onset asth-ma.22 Despite evidence that obesity and asthmatend to co-occur, relatively few studies23–25 haveexamined whether BMI is actually associated withworse asthma severity, or the extent to whichpatients with higher BMIs exhibit worse control oftheir asthma or worse asthma-related quality oflife. Given recent treatment guidelines emphasiz-ing improved asthma control and quality of life,26

and the fact that obesity may be considered adisorder related to poor self-management,27 itseems important to investigate the potentialimpact of increasing BMI on these behaviorallyinfluenced variables.

The purpose of the present study was threefold:(1) to estimate the prevalence of overweight andobesity in a Canadian tertiary care population ofadult asthma patients; (2) to evaluate associationsbetween BMI and asthma severity; and (3) toevaluate associations between BMI and levels ofasthma control and asthma-related quality of life.

Material and methods

Study participants

Participants were recruited from consecutive pa-tients presenting to the asthma clinic of Hopital duSacre-Coeur de Montreal from June 2003 to April2004. This study reflects data collected as part of alarger study evaluating the prevalence and impactof psychiatric disorders among adult asthmatics.28

Patients were eligible if they had documentedphysician-diagnosed asthma, were between theages of 18 and 75, and fluent in English or French.A total of 990 patients presented to the asthmaclinic, of which 865 (87%) were screened forinclusion in the study (the remaining 125 hadinsufficient medical information with which toconduct pre-screening). A total of 442 patientswere excluded for failing to meet eligibility criteria(see Fig. 1), resulting in 423 eligible patients whowere contacted to participate in the study. Only 34patients declined to participate, which yielded a

ARTICLE IN PRESS

Presented to Clinic

990

Screened for eligibility

865 (87%)

Eligible participants

423 (49%)

Consenting participants

389 (92%)

Final sample

382

Excluded

442 (51%)

- Co-morbid disease = 180

- Unconfirmed asthma =136

- Age criteria =107

Refused to participate

34

Excluded from analyses

7

- Missing data = 6

- BMI > 50 (Kg/m2) = 1

Figure 1 Flow chart of patient screening, eligibility andparticipation.

K.L. Lavoie et al.650

final sample of 389 patients (92% participationrate). Six patients were excluded from the analysesdue to incomplete or missing data, and one patientwas excluded for having an outlying BMI of 450 kg/m2, yielding a final sample of 382 patients. Thisproject was approved by the Ethics Committee ofHopital du Sacre-Cœur de Montreal.

Study design

Patients were screened on the day of their asthmaclinic visit to verify eligibility. Asthma diagnosis wasconfirmed by chart review and verified by thetreating physician. Participants underwent a demo-graphic and medical history interview, includingasking patients to report whether a doctor haddiagnosed them with various medical conditions(e.g., hypertension/high blood pressure; hyperch-olesterolemia/high cholesterol, and diabetes). Pa-tients were also asked to report height and weight.All height and weight variables were converted intokg and m to calculate BMI. Patients completed theAsthma Control (ACQ) and Asthma Quality of Life(AQLQ) Questionnaires and underwent objectivestandard pulmonary testing (spirometry) in therespiratory laboratory as part of their regular

physician visit. Medical information and medicationstatus and dosage were verified by chart review.

Obesity and asthma severity assessment

BMI (in kg/m2) was calculated from patients’ self-reported height and weight. Because this study waspart of a larger study that involved extensive andtime-consuming psychological assessments, werelied upon self-reported height and weight, whichhas been shown to be almost identical to and highlycorrelated with measured values.25 The interna-tional standard definition of obesity was used.9,10

Patients were classified as normal if BMI was o25;overweight if their BMI was X25o30; and obese ifBMI was X30. Only one patient had an excessivelyhigh BMI (X50) and was excluded. Asthma severitywas defined according to the updated 2002 GINAguidelines26 which categorize asthma severity intofour categories (intermittent, mild persistent,moderate persistent, and severe persistent).

Asthma control and quality of life assessment

Participants completed the ACQ30 and the AQLQ.31

The ACQ evaluates asthma control by askingrespondents to recall their symptoms and b2-agonsist use in the last week, plus one questionassessing spirometry results (FEV1, % Pred.). TheACQ contains 7 items rated on a 7-point scale (0 ¼good control, 6 ¼ poor control). The AQLQevaluates asthma quality of life across four do-mains: activity limitation, symptoms, emotionaldistress, and environmental stimuli. It contains 32items rated on a 7-point scale (1 ¼ maximalimpairment, 7 ¼ no impairment). Both question-naires have demonstrated very good measurementproperties, including high intra-class correlationcoefficients between 0.90 and 0.95 and goodconstruct, cross-sectional and longitudinal valid-ity,30–33 and have been validated in QuebecFrench.34

Pulmonary function testing

All pulmonary testing using standard spirometrywas conducted in the respiratory laboratory by atrained respiratory technician on the day of thepatient’s clinic visit. Rescue medication was with-held for at least 4 h before pulmonary functiontests. FEV1 and FVC were assessed before admin-istration of 200mcg salbutamol using a metered-dose inhaler or 500 mg terbutaline using a Turbu-halers. Predicted values of FEV1 and FVC werecalculated from reference values for patients less

ARTICLE IN PRESS

BMI, asthma severity, asthma control and quality of life 651

than35 and greater36 than 70 year, respectively,yielding % predicted FEV1 and % predicted FEV1/FVC.

Data analyses strategy

In order to maximize statistical power and due tothe increased potential for bias when categories(e.g., weight categories) are used in place ofcontinuous data,37 relationships between BMI (asa continuous variable) and sociodemographic vari-ables, medical characteristics, asthma character-istics, and pulmonary function were analyzed usinggeneral linear models (GLM). The relationshipbetween BMI and asthma severity was also analyzedusing GLM controlling for age, sex, and asthmaduration. Relationships between BMI and ACQ andAQLQ scores were analyzed using GLMs controllingfor age, sex, and asthma severity. Though statistics(F’s) will reflect continuous BMI results, all data ispresented as a function of BMI category (normal,overweight, obese) in order to more easily inter-pret the pattern of results. All tests were two-sidedand significance was set at .05. Data analysis wasperformed using SAS v8.2 (SAS Institute, Cary NC).

Results

Sample characteristics

A total of 382 asthma patients participated in thepresent study. Participants were 60% female andhad a mean age of 49 (714.2) years. The mean BMIfor the entire sample was 27 kg/m2 (SD ¼ 5.1). Based

Table 1 Patient demographics and medical characterist

% (n) Normal BMIo25(n ¼ 139)

OverweighBMIX25o(n ¼ 149)

Age� (yr) 47 (15) 49 (14)Sex (female) 70 (97) 48 (72)Caucasian 95 (131) 92 (137)Years of education� 13.8 (3.2) 12.9 (3.7)Employed 62 (83) 66 (98)Married/cohabitating 61 (81) 70 (104)Hypertension 17 (23) 19 (28)Diabetes 2 (3) 5 (8)Hypercholesterolemia 14 (19) 17 (25)Ever smoked 46 (63) 54 (81)Current smoker 15 (21) 8 (12)

�Value is M7SD.

on the international BMI classification, a total of139 (36%) patients had a normal BMI; 149 (39%)patients were overweight; and 94 (25%) patientswere obese.

Sociodemographics and comorbid medicalconditions

Sociodemographic characteristics and comorbidmedical conditions presented as a function of BMIcategory are presented in Table 1. There weresignificant associations between BMI and age(F ¼ 8:42, Po:01) and BMI and hypertension(F ¼ 14:67, Po:0001). Patients with a higher BMIwere more likely to be older and have hypertensionthan patients with a lower BMI. No other significantrelationships were observed.

Asthma characteristics and pulmonaryfunction

Asthma characteristics and pulmonary functionpresented as a function of BMI category arepresented in Table 2. There was a significantassociation between BMI and the proportion ofpatients taking long-acting bronchodilators(F ¼ 4:44, P ¼ :04) and a significant negativerelationship between BMI and FVC (F ¼ 6:79,Po:01). Patients with a higher BMI were morelikely to be taking long-acting bronchodilators andexhibit a lower FVC during pulmonary functiontesting. No other significant relationships wereobserved.

ics as a function of BMI category.

t30

Obese BMIX30(n ¼ 94)

F P

51 (12) 8.42 o.0168 (64) 0.07 .7993 (87) 0.19 .66

12.5 (4.3) 2.04 .1557 (51) 0.01 .9167 (60) 1.76 .1930 (28) 14.67 o.00019 (8) 3.31 .07

23 (22) 0.54 .4655 (52) 2.18 .1413 (12) 1.21 .27

ARTICLE IN PRESS

Table 2 Asthma characteristics and pulmonary function as a function of BMI category.

Normal BMIo25(n ¼ 139)

OverweightBMIX25o30(n ¼ 149)

Obese BMIX30(n ¼ 94)

F P

Asthma severity, M (SD) 3.2 (.7) 3.2 (.6) 3.3 (.6) 3.60 .06Asthma duration, yrs, M (SD) 19 (15) 20 (15) 20 (16) 0.86 .35Atopy, % (n) 65 (89) 64 (95) 63 (59) 0.16 .69Daily ICS dose� M (SD) 641 (549) 703 (547) 799 (558) 3.50 .06# SAB last week, M (SD) 11 (20) 8 (15) 11 (16) 1.06 .30SAB, % (n) 98 (135) 97 (144) 99 (93) 0.06 .22LAB, % (n) 63 (87) 74 (110) 72 (66) 4.44 .03ICS, % (n) 91 (125) 94 (140) 95 (89) 2.96 .09Oral CS, % (n) 13 (17) 10 (15) 16 (15) 0.00 .98Anti-leukotrienes, % (n) 15 (21) 14 (20) 14 (13) 0.14 .71Anti-Rhinitics, % (n) 17 (23) 24 (35) 20 (18) 0.25 .62FEV1, % pred. M (SD)y 79 (20) 79 (21) 75 (19) 2.08 .15FVC, % pred. M (SD)y 90 (18) 91 (18) 84 (18) 6.79 .01FEV1/FVC, % pred. M (SD)y 72 (12) 70 (11) 73 (10) 0.33 .56

ICS ¼ inhaled corticosteroid; SAB ¼ short acting bronchodilator; LAB ¼ long acting bronchodilator; FEV ¼ forced expiratoryvolume; FVC ¼ forced vital capacity.�Flucatisone propionate equivalent (mg).yPre-bronchodilator.

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Intermittent Mild Moderate Severe

Normal Overweight Obese

Figure 2 Percentage of patients in each BMI categorypresented as a function of asthma severity category.

K.L. Lavoie et al.652

Asthma severity

Asthma severity presented as a function of BMIcategory is presented in Fig. 2. Although there wasa non-significant trend for asthma severity to beworse with increasing BMI (F ¼ 3:60, P ¼ :06, seeTable 2), when age, sex and asthma duration werecontrolled for there was no significant relationship(F ¼ 1:60, P ¼ :21).

BMI� sex interactionDue to previous reports suggesting stronger asso-ciations between increasing BMI and asthma21,38

and asthma severity39 in women relative to men,we conducted supplementary GLM analyses of theinteraction between BMI and sex (controlling forage and asthma duration). The analyses revealedno significant BMI� sex interaction for asthmaseverity (F ¼ 0:25, P ¼ :63).

Asthma control and quality of life

Scores on the ACQ and AQLQ scales presented as afunction of BMI category are presented in Table 3.There were significant associations between BMIand total ACQ score (F ¼ 5:96, Po:01), indicatingthat patients with a higher BMI had higher scores onthe ACQ, independent of age, sex, and asthmaseverity. There were also significant associationsbetween BMI and several individual items on theACQ. Patients with a higher BMI reported signifi-cantly greater waking symptoms (F ¼ 3:81, Po:05),significantly greater limitations in their dailyactivities (F ¼ 16:08, Po:001); significantly moreshortness of breath (F ¼ 11:48, Po:001); signifi-cantly more wheezing (F ¼ 6:76, Po:01); andsignificantly greater bronchodilator use (F ¼ 3:64,Po:05). There was also a significant association

ARTICLE IN PRESS

Table 3 Asthma control and quality of life as a function of BMI category, controlling for age, sex, and asthmaseverity.

M (7SEM) Normal BMIo25(n ¼ 139)

OverweightBMIX25o30(n ¼ 149)

Obese BMIX30(n ¼ 94)

F P

ACQ (total) 1.637.09 1.607.08 1.937.10 5.96 .01Q 1: Nocturnal waking 1.137.13 0.907.12 1.267.15 1.24 .26Q 2: Waking symptoms 1.417.13 1.417.12 1.647.15 3.81 .05Q 3: Activity limitations 1.117.12 1.237.11 1.797.14 16.08 o.001Q 4: Shortness of breath 1.847.13 1.827.12 2.347.15 11.48 .001Q 5: Wheezing 1.607.12 1.557.12 2.067.15 6.76 .01Q 6: Bronchodilator use 1.177.11 1.027.11 1.537.13 3.94 .05Q 7: % FEV1 2.837.13 2.707.12 2.847.16 0.23 .63

AQLQ (total) 5.187.11 5.177.11 4.847.13 6.30 .01Symptoms 5.167.12 5.227.11 5.007.14 2.16 .14Emotions 5.507.12 5.467.12 5.327.15 2.13 .14Environmental stimuli 5.067.14 4.917.13 4.407.17 13.38 o.001Activity limitation 5.097.13 5.087.12 4.637.15 8.09 .005

ACQ ¼ Asthma Control Questionnaire; AQLQ ¼ Asthma Quality of Life Questionnaire.

BMI, asthma severity, asthma control and quality of life 653

between BMI and total AQLQ score (F ¼ 6:30,Po:01), indicating that patients with a higher BMIhad lower scores on the AQLQ, independent of age,sex, and asthma severity. There was also asignificant association between BMI and the envir-onment subscale (F ¼ 13:38, Po:001) and theactivities subscale (F ¼ 8:09, Po:01) of the AQLQ,indicating that patients with a higher BMI had worseasthma-related quality of life by being morelimited in their activities and experiencing greatersymptoms in response to environmental triggers.

BMI� sex interactionControlling for sex in our analyses may haveobscured potential sex differences in the observedassociations between increasing BMI and ACQ andAQLQ scores. We therefore conducted a series ofGLM’s to assess any potential BMI� sex interactioneffects on ACQ and AQLQ scores, controlling for ageand asthma severity. The analyses revealed nosignificant BMI� sex interactions for total ACQscore or for any of the individual item scores (allP’s4.05; see Fig. 3a for scatter plot of total ACQscores). However, the analyses did reveal a sig-nificant pattern of BMI� sex interactions for AQLQscores, including total AQLQ score (F ¼ 5:78,Po:05; see Fig. 3b), and for the activity limitation(F ¼ 8:97, Po:01) and environment (F ¼ 4:02,Po:05) subscales, indicating that with increasingBMI, there is a sharper decline in asthma-relatedquality of life (including activity limitations and inrelation to environmental triggers) in men com-pared to women.

Discussion

The results of the present study indicate a high rateof both overweight and obesity among adultasthmatics seen in a Canadian tertiary care asthmaclinic, and that having a higher BMI is associatedwith worse asthma control and quality of lifeindependent of age, sex and asthma severity. Thissuggests that higher BMIs are associated withclinically relevant asthma morbidity, and thatexcess body fat may complicate both asthmacontrol and quality of life independently of factorstypically related to these variables (i.e., age,asthma severity). We found no significant associa-tion between BMI and asthma severity, aftercontrolling for age and sex. This finding is incontrast to at least two studies23,25 who did findsignificant associations between increasing BMI andasthma severity. However, our study differed fromthese two in at least two important ways: Acker-man et al.23 failed to control for any potentialconfounding variables, which may have at leastpartially accounted for their results; and Vassaroet al.25 measured asthma severity using a compo-site score rather than an established classificationsystem (i.e., 1997 NHLBI or 2002 GINA guidelines),which may also have at least partially accountedfor our discrepant findings.

The present findings did not reveal any significantsex differences in the relationship between BMI andasthma severity, which is consistent with theresults of at least two studies.24,40 However, ourfindings are contrary to those reported by theENFUMOSA Study Group,39 who did observe sex

ARTICLE IN PRESS

0

1

2

3

4

5

6

15 20 25 30 35 40 45 50BMI (kg/m2)

AC

Q t

ota

l

Women

Men

1

2

3

4

5

6

7

15 20 25 30 35 40 45 50

BMI (kg/m2)

AQ

LQ

to

tal

Women

Men

Men are represented with a dotted (----) line and women with a solid ( ) lineBMI × gender interaction (F = 5.78, p < .05)

Men are represented with a dotted (----) line and women with a solid ( ) lineBMI × gender interaction (F = 0.01, p = .93) (a)

(b)

Figure 3 (a) Scatter plot of the non-significant BMI� sex interaction for total ACQ score. (b) Scatter plot of thesignificant BMI� sex interaction for total AQLQ score.

K.L. Lavoie et al.654

differences in the relationship between BMI andasthma severity. This may be due to differences insampling procedures and/or differences in the wayasthma severity was classified and defined. Weincluded a consecutive sample of eligible patients,which may be less subject to selection bias thanconvenience sampling that appears to have been

used in the ENFUMOSA study. We also classifiedasthma severity according to established criteria(2002 GINA criteria), which differs from theENFUMOSA study who defined severe asthmaaccording to inhaled corticosteroid treatmentlevels and asthma exacerbations in the last year.Given the potential importance of identifying sex

ARTICLE IN PRESS

BMI, asthma severity, asthma control and quality of life 655

differences in the relationship between BMI andasthma severity, future studies should use consis-tent definitions of asthma severity in order tofacilitate comparison across studies.

Though we found no sex differences in therelationship between BMI and asthma control, wedid observe interesting sex differences in BMI–AQLQassociations, with men experiencing a sharperdecline in asthma-related quality of life comparedto women with increasing BMI. To our knowledge,this is the first study to report such an association,and suggests that asthmatic men, rather thanasthmatic women, may be at greater risk ofworsening asthma-related quality of life withincreased body weight. Though this study was notdesigned to investigate why such sex differencesmay exist, we can speculate as to why they mayhave occurred. An examination of the AQLQsubscales that showed sex differences indicatesthat men appear to report worse asthma-relatedquality of life on the two subscales (activitylimitation and environmental stimuli) that mostreflect decreases in the functional aspect of qualityof life as a result of asthma. In general, men aremore active than women.41 As such, the loss offunctional capacity with increasing BMI may havegreater meaning for men than women, which mayin part account for our present findings. Futureresearch is needed to explore these hypotheses.

A total of 64% of this sample of asthma patientswere either overweight or obese, which is nearly15% more than the national (Canadian) rates ofoverweight and obesity.21 Moreover, one out of four(25%) of asthmatics were considered obese, a ratethat exceeds the national average of 15% by overtwo-thirds.11,21 This suggests that obesity may bemore common among adult tertiary care asthmaticsthan in the general population, a finding consistentwith studies from other industrialized na-tions.8,19,42 To our knowledge, this is one of onlytwo studies to report rates of overweight andobesity in a Canadian sample of adult asthmatics,with the other community based study reporting asimilar (though somewhat reduced) rate of obesity(20%).21 Taken together, these studies indicate arobust relationship between BMI and asthma in bothcommunity and tertiary care settings in Canada.

Previous reports have linked either higher BMI orobesity to increased prevalence of asthma symp-toms such as wheezing43–45 and shortness ofbreath.44,45 The present study strengthens thisdata by assessing a range of asthma symptomatol-ogy, including measures of symptoms, medicationuse, and functional status, and demonstratingthese associations independent of disease severity.The AQLQ data also reveals a pattern consistent

with greater functional impairment in asthmapatients with higher BMIs, both in terms of beingmore limited in their activities and experiencinggreater symptoms in response to environmentaltriggers. Though our findings reveal a potentiallystronger BMI–AQLQ relationship in men, our mainanalyses revealed significant BMI–AQLQ associationsindependent of sex. Although it may be argued thatpatients with higher BMIs may be more restricted intheir daily activities due to their excess weight,there are at least two points which make thisconclusion unlikely. First, patients are specificallyquestioned about activity limitations due to theirasthma. Second, the relationships observed arebetween BMI in its continuous form, and eachvariable of interest. As such, the relationshipbetween activity limitation and BMI is true evenfor patients who fall within the normal range ofBMI, but who have a relatively higher BMI thanothers in their category. Therefore, having a higherBMI is associated with greater activity limitation inpatients at all levels of BMI categories and for thosewho may not necessarily be obese.

Though this study was not designed to evaluatewhy asthmatics with increasing BMI may have worseasthma control and asthma-related quality of life,we can hypothesize as to why these associationswere observed. It is known that obese patients tendto experience wheezing and breathlessness due totheir excess weight.45,46 It is therefore possiblethat obese patients may have falsely attributedweight-related respiratory symptoms to asthma,causing increased use of their bronchodilator.44

However, the fact we found evidence of worseasthma control across the entire range of BMI (andnot exclusively in patients who were obese) makesthis explanation unlikely. Thompson et al.24 alsoreported no evidence of less ‘‘real’’ asthma inobese versus non-obese asthmatics, and is consis-tent with our findings of comparable pulmonaryfunction (i.e., FEV1, FEV1/FVC, % predicted) (withthe exception of FVC which likely reflects mechan-ical limitations in VC), across all levels of BMI.Given that being overweight or obese suggestschronic difficulties with self-management of dietand exercise behavior, an alternative explanationmay be that asthmatics with increasing BMI mayhave worse asthma control due to behavioralfactors associated with poor asthma self-manage-ment. There is evidence to suggest that increases inBMI are associated with both delay and avoidanceof health care seeking behavior,47 which may haveimplications for symptom monitoring and ensuringadequate symptom control. If optimal asthmacontrol depends on regular contact with treat-ing physicians, targeting overweight and obese

ARTICLE IN PRESS

K.L. Lavoie et al.656

asthmatics for behavioral interventions and possi-bly healthcare workers to improve attitudes to-wards overweight and obese patients, representimportant goals for clinical research. There areencouraging results from at least two studiesshowing improvements in respiratory function,airway obstruction, and peak expiratory flowvariability following weight reduction in obesepatients with asthma.48,49 Future studies examiningthe effects of weight loss on asthma should includemeasures of asthma control and functional capacity(quality of life) to determine whether weight lossinterventions can also impact behavioral variables.

There are some limitations to the present study.One involves the use of self-reported height andweight for the calculation of BMI. However, datafrom the National Health and Nutrition ExaminationSurvey (NHANES-III) found that self-reported heightand weight does not differ substantially frommeasured values.29 In fact, BMI values derived frommeasured and self-reported height and weight arehighly correlated (r ¼ 0:8920:97, P’so.001).25

Moreover, the sensitivity and specificity of self-reported to predict measured values of height andweight are excellent for adults under the age of 80,ranging from 86–94% (sensitivity) to 96–99% (speci-ficity), with a tendency for self-reported measuresto be underestimates. As we excluded patients overthe age of 75, we are quite confident in ourestimates of BMI. Another potential limitation ofthis study involves the nature of the population westudied. We sampled asthmatics followed at atertiary care hospital whose asthma may be some-what more severe than patients followed by primarycare physicians or community samples. Therefore,results may not generalize to non-tertiary-careasthma patients. We also did not assess hormonalvariables which may be important mediators ofBMI–asthma relationships in women.25 Finally, werestricted our measures of excess body weight toBMI, however other measures, such as centraladiposity or abdominal obesity,50,51 may also beimportant. Future studies assessing the impact ofincreasing body fat on asthma should include abroader range of measures such as waist–hip ratio,waist/abdominal circumference, and direct mea-sures of total body fat in order to determine therelative importance of each measure.

In conclusion, our findings suggest that higher BMImay have important implications for asthma symp-toms, medication use, and functional impairment,and that men may experience a sharper decline inasthma-related quality of life with increasing BMI.Though previous studies have found associationsbetween obesity and increased asthma symptoma-tology, no study has specifically evaluated whether

asthmatics with higher BMIs demonstrate worseasthma control and quality of life irrespective ofBMI ‘‘category’’. Results suggest that patients neednot be classified as ‘‘obese’’ for excess weight andbody fat to compromise several important asthmavariables. Findings identify BMI and obesity asimportant behavioral factors in asthma controland suggest important avenues for asthma manage-ment and control initiatives. Though there isevidence of improved lung function followingweight reduction in obese asthmatics, futurestudies should assess a broader range of bodyweight measures and include measures of asthmacontrol and quality of life to determine whetherweight loss interventions can impact multiplemeasures of body weight as well as behavioralvariables.

Acknowledgements

The authors thank Dr. Catherine Lemiere , Dr. Jean-Luc Malo, and Mr. Guillaume Lacoste for theirinvaluable assistance with patient recruitment. Thisproject was supported by a fellowship grant fromthe Canadian Institutes of Health Research (KLL) anda research grant from the Fondation de l’Hopital duSacre-Coeur de Montreal (Auger Foundation).

References

1. Statistics Canada, Cansim, tables 104-0001 & 105-001,Catalogue no. 82-221-XIE, 2004.

2. CDC, National Center for Health Statistics. Asthma pre-valence, health care use and mortality, 2000–2001. Avail-able at http://www.cdc.gov/nchs/products/pubs/pubd/hestats/asthma/asthma.htm.

3. National Center for Health Statistics. Current estimatesfrom the National Health Interview Survey. Vital Health Stat1990;10(194):1997.

4. National Heart, Lung and Blood Institute. Data fact sheet.Asthma Statistics. Bethesda, MD: National Institutes ofHealth, Public Health Service; 1999.

5. Weiss KB, Gergen PJ, Hodgson TA. An economic evaluation ofasthma in the United States. N Eng J Med 1992;326:862–6.

6. Krahn MD, Berka C, Langlois P, et al. Direct and indirect costsof asthma in Canada, 1990. CMAJ 1996;154:821–31.

7. Chapman KR, Ernst P, Grenville A, Dewland P, Zimmerman S.Control of asthma in Canada: failure to achieve guidelinetargets. Can Respir J 2001;8(Suppl A):35A–40A.

8. Flegel KM, Caroll MD, Kuczmarksi RJ, et al. Overweight andobesity in the United States: prevalence and trends,1960–1994. Int J Obes Relat Metab Disord 1998;22:39–47.

9. World Health Organization. Obesity: Preventing and Mana-ging the Global Epidemic, Report of a WHO Consultation onObesity, Geneva, Geneva, June 3–5, 1997. Geneva, Switzer-land: WHO; 1998.

10. National Institutes of Health (NIH). NHLBI: Clinical guide-lines on the identification, evaluation and treatment of

ARTICLE IN PRESS

BMI, asthma severity, asthma control and quality of life 657

overweight and obesity in adults: the evidence report. NIHreport 98-4083, 1998.

11. Katzmarzyk PT. The Canadian obesity epidemic, 1985–1998.CMAJ 2002;166:1039–40.

12. Wolf AM, Colditz CA. Current estimates of the economic costof obesity it the United States. Obes Res 1998;6:97–106.

13. Wellman NS, Friedberg B. Causes and consequences of adultobesity: health, social and economic impacts in the UnitedStates. Asia Pac J Clin Nutr 2002;11:S705–9.

14. Birmingham CL, Muller JL, Palepu A, et al. The cost ofobesity in Canada. AMAJ 1999;160:483–8.

15. Pi-Sunyer FX. Medical hazards of obesity. Ann Int Med1993;119:655–60.

16. Gennuso J, Epstein LH, Paluch RA, et al. The relationshipbetween asthma and obesity in urban minority children andadolescents. Arch Pediatr Adolesc Med 1998;152:1197–200.

17. Figuerosa-Munoz JI, Chinn S, Rona RJ. Association betweenobesity and asthma in 4–11 year old children in the UK.Thorax 2001;56:133–7.

18. Epstein LH, Wu YW, Paluch RA, et al. Asthma and maternalbody mass index are related to pediatric body mass indexand obesity: results from the Third National Health andNutrition Examination Survey. Obes Res 2000;8:575–81.

19. Luder E, Ehrlich RI, Lou WYW, et al. Body mass index and therisk of asthma in adults. Resp Med 2004;98:29–37.

20. Sunghye K, Camargo CA. Sex–race differences in therelationship between obesity and asthma: The behavioralrisk factor surveillance system, 2000. Ann Epidemiol2003;13:666–73.

21. Chen Y, Dales R, Tang M, et al. Obesity may increase theincidence of asthma in women but not men: longitudinalobservations from the Canadian National Population HealthSurveys. Am J Epidemiol 2002;155:191–7.

22. Camargo CA, Weiss ST, Zhang S, et al. Prospective study ofbody mass index, weight change, and risk of adult-onsetasthma in women. Arch Int Med 1999;159:2582–8.

23. Ackerman MJH, Calacanis CM, Madsen MK. Relationshipbetween asthma severity and obesity. J Asthma 2004;41:521–6.

24. Conley-Thomson C, Clark S, Carmago CA. Body mass indexand asthma severity among adults presenting to theemergency department. Chest 2003;124:795–802.

25. Vassaro R, Siroux V, Maccario J, et al. Asthma severity isassociated with body mass index and early menarche inwomen. Am J Respir Crit Care Med 2005;171:334–9.

26. US Department of Health and Human Services, NationalInstitutes of Health, National Heart, Lung and BloodInstitute. Global initiative for asthma: global strategy forasthma management and prevention. Washington, DC:US Government Printing Office; 2002 (NIH publicationno. 02-3659).

27. National Institutes of Health, National Heart, Lung, andBlood Institute Obesity Education Initiative. Clinical guide-lines on the identification, evaluation and treatment ofoverweight and obesity in adults. Bethesda, MD: NationalInstitutes of Health; 1998.

28. Lavoie KL, Cartier A, Labrecque M, et al. Are psychiatricdisorders associated with worse asthma control and qualityof life in asthma patients? Respir Med 2005;99:1249–57.

29. Kuczmarski MF, Kuczmarski RJ, Najjar M. Effects of age onvalidity of self-reported height, weight, and body massindex: Findings from the third National Health and NutritionExamination Survey, 1988–1994. Diet Assoc 2001;101:28–34.

30. Juniper EF, O’Byrne PM, Guyatt GH, et al. Development andvalidation of a questionnaire to measure asthma control. EurRespir J 1999;14:902–7.

31. Juniper EF, Guyatt GH, Ferrie PJ, et al. Measuring quality oflife in asthma. Am Rev Respir Dis 1993;147:832–8.

32. Juniper EF, O’Byrne PM, Ferrie PJ, et al. Measuring asthmacontrol: Clinic questionnaire or daily diary? Am J Respir CritCare Med 2000;162:1330–4.

33. Leidy NK, Coughlin C. Psychometric performance of theAsthma Quality of Life Questionnaire in a US sample. QualLife Res 1998;7:127–34.

34. Juniper, EF: http://www.qoltech.co.uk/index.htm.35. Knudson RJ, Lebowitz MD, Holberg CJ, et al. Changes in the

normal maximal expiratory flow curve with growth andaging. Am Rev Resp Dis 1983;127:725–34.

36. Enright PL, Kronmal RA, Higgins M, et al. Spirometryreference values for women and men 65 to 85 years ofage. Am Rev Respir Dis 1993;147:125–33.

37. MacCallum RC, Zhang S, Preacher KJ, et al. On the practiceof dichotomization of quantitative variables. Psychol Meth-ods 2002;7:19–40.

38. Beckett WS, Jacobs Jr. DR, Yu X, et al. Asthma is associatedwith weight gain in females but not males, independent ofphysical activity. Am J Respir Crit Care Med 2001;164:2045–50.

39. ENFUMOSA Study Group. The ENFUMOSA cross-sectionalEuropean multicentre study of the clinical phenotype ofchronic severe asthma. Eur Respir J 2003;22:470–7.

40. Chinn S, Jarvis P, Burney P. Relation of bronchial respon-siveness to body mass index in the ECRHS. Thorax2002;57:1028–33.

41. Brown DR, Yore MM, Ham SA, Macera CA. Physical activityamong adults 4or ¼ 50 yr with and without disabilities,BRFSS. Med Sci Sports Exerc 2005;37:620–9.

42. Gueurra S, Sherrill DL, Bobadilla A, et al. The relation ofbody mass index to asthma, chronic bronchitis, andemphysema. Chest 2002;122:1256–63.

43. Schwartz J, Gold D, Dockery DW, et al. Predictors of asthmaand persistent wheeze in a national sample of children in theUnited States: associations with social class, perinatalevents, and race. Am Rev Respir Dis 1990;142:555–62.

44. Schacter LM, Salome CM, Peak JK, et al. Obesity is a riskfactor for asthma and wheeze but not airway hyperrespon-siveness. Thorax 2001;56:4–8.

45. Jarvis D, Chinn S, Potts J, et al. On behalf of the EuropeanCommunity Respiratory Health Survey. Association of bodymass index with respiratory symptoms and atopy: resultsfrom the European Community Health Survey. Clin ExperAllergy 2002;32:831–7.

46. Sin DD, Jones RL, Man SF. Obesity is a risk factor for dyspneabut not for airflow obstruction. Arch Int Med 2002;162:1477–81.

47. Drury CA, Louis M. Exploring the association between bodyweight, stigma of obesity, and health care avoidance. J AmAcad Nurse Pract 2002;14:554–61.

48. Hakala K, Stenius-Aarniala B, Sovijarvi A. Effects of weightloss on peak flow variability, airways obstruction, and lungvolumes in obese patients with asthma. Chest 2000;118:1315–21.

49. Aaron SD, Fergusson D, Dent R, et al. Effect of weightreduction on respiratory function and airway reactivity inobese women. Chest 2004;125:2046–52.

50. Despres JP, Lemieux I, Prud-homme D. Treatment of obesity:need to focus on high risk abdominally obese patients. BMJ2001;322:716–20.

51. Nystrom Kronander U, Falkenburg M, Zetterstrom O.Prevalence and incidence of asthma related to waistcircumference and BMI in a Swedish community sample.Respir Med 2004:1108–16.