| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Guest Access | Sign In via User Name/Password
|
|
|
|||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Arizona Respiratory Center, University of Arizona, College of Medicine, Tucson, AZ.
Correspondence to: Robert A Barbee, MD, FCCP, Arizona Respiratory Center, University of Arizona, 1501 N. Campbell Ave, PO Box 245030, Tucson, AZ 85724-5030
 |
Abstract |
|---|
 TOP Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Objectives: To determine whether a relation of body mass index (BMI) to asthma, chronic bronchitis (CB), or emphysema exists (analysis 1), and, if so, whether the association between obesity and asthma is modified by gender (analysis 2).
Design: Nested case-control study from the longitudinal cohort of the Tucson Epidemiologic Study of Airways Obstructive Diseases.
Patients: Analysis 1: physician-confirmed incident cases of asthma (n = 102), CB (n = 299), or emphysema (n = 72) who denied any prior AODs. Analysis 2: all 169 incident cases of asthma, regardless of any previous AODs, stratified by gender and by other potential effect modifiers. In both analyses, we selected only subjects at least 20 years old who had weight and height measured during the study.
Measurements: BMI and other risk factors were assessed prior to the onset of the AOD (cases) or prior to the last completed survey (control subjects).
Results: A diagnosis of emphysema was significantly associated with a BMI < 18.5 (odds ratio [OR], 2.97; 95% confidence interval [CI], 1.33 to 6.68, when compared to healthy control subjects). A BMI 
28 increased the risk of receiving a diagnosis of asthma (OR, 2.10; 95% CI, 1.31 to 3.36) and CB (OR, 1.80; 95% CI, 1.32 to 2.46). About 30% of the patients with asthma and 25% of the patients with CB (vs 16% of the control subjects, p < 0.001) were preobese or obese, regardless whether BMI was assessed before the diagnosis or before the onset of respiratory symptoms. The relation of elevated BMI to asthma was significant only among women.
Conclusions: Patients with emphysema are more likely to be underweight, and patients with CB are more likely to be obese. However, the temporal relationship between abnormal BMI and the onset of COPD is uncertain. Preobese and obese women are at increased risk of acquiring asthma. This relation, particularly if it is causal, has potentially relevant public health implications.
Key Words: asthma • body mass index • chronic bronchitis • COPD • emphysema • obesity
|
Introduction |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Chen et al1
found that a BMI < 20 among male subjects and a BMI 28 among female subjects were associated with an increased prevalence of COPD (CB or emphysema). However, the potential confounding effect of comorbid asthma was not addressed and, within the COPD group, no distinction was made between a diagnosis of CB and emphysema.
The relationship to an abnormal BMI has been more extensively studied for asthma than COPD. Higher levels of BMI and higher proportions of obese subjects have been found among patients with asthma than control subjects, both in pediatric2 3 4 and adult5 6 7 8 populations. Several studies2 3 4 5 6 have reported this association to be stronger among female than male subjects.
To our knowledge, only one study9 used longitudinal data to address the relationship between obesity and asthma among adults. In that study, within a prospective cohort of 85,911 female US nurses, Camargo et al9 found that the BMI, as measured in 1991, strongly correlated with the risk of adult-onset asthma developing in the following 4 years. Moreover, women who gained weight after the age of 18 years were more likely to acquire asthma during the same follow-up period.
Based on this previous evidence, we aimed to test, in a longitudinal community population, three working hypotheses: (1) an abnormally low or high BMI is associated with AODs among adults, (2) the relation of abnormal BMI to asthma and CB does not differ when weight is assessed before the onset of respiratory symptoms rather than before the diagnosis, and (3) the association between abnormal BMI and asthma does differ between female and male subjects.
|
Materials and Methods |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Between 1972 and 1992, participating subjects were monitored in 12 periodic surveys, obtained approximately 1.5 to 2 years apart. During each survey, they completed a standardized questionnaire providing information on respiratory symptoms, pulmonary diseases, occupation, smoking history, as well as histories of other diseases. Except during survey 4, participants also had their weight and height measured by the study nurses with standardized equipment.
BMI was computed as weight divided by the square of height, and was grouped into four categories: underweight (BMI < 18.5), normal (BMI 18.5 and < 25), mild overweight (BMI 25 and < 28), and preobese/obese (BMI 28). Allergy skin-prick tests for the most common allergens in the Tucson area were performed up to three times for each subject during the study period. A subject was considered atopic if a reaction at least 2 mm greater than the size of the control was recorded for at least one of the tested allergens. Blood samples were obtained and serum IgE levels were measured up to three times for each subject at least 6 years old during the study. The paper radioimmunosorbent test (PRIST; Pharmacia Diagnostics; Piscataway, NJ) method was used for the measurement of total serum IgE.
Analysis 1
Among subjects who had weight and height measured, we selected subjects at least 20 years old without any previous AODs and with a new physician-confirmed diagnosis of asthma (n = 102), CB (n = 299), or emphysema (n = 72) during the study period (incident cases). In order to code our cases as physician confirmed, we required them not only to report the presence of asthma, CB, or emphysema, but also to have been seen, diagnosed, or treated for such a disease by a doctor.
Control subjects were represented by 1,475 subjects who reported no AOD diagnosis and no respiratory symptoms (shortness of breath with wheezing, or cough and phlegm for at least 3 months in 2 consecutive years) both at enrollment and during the entire study period. Control subjects were also required to have completed at least two surveys, to have weight and height measured during the study, and be 20 years old at the last completed survey.
For cases, the presence of potential risk factors was assessed at the last nonmissing survey preceding the first report of asthma, CB, or emphysema. Similarly, among the control subjects, potential risk factors were assessed prior to the last completed survey.
As it could be argued that subjects with AOD are more likely to avoid exercise and gain weight because of their respiratory symptoms, in a limited number of cases we also assessed BMI before the onset of respiratory symptoms (shortness of breath with wheezing for asthmatics, and chronic cough and phlegm for patients with CB). These were cases in which the diagnosis of AOD had been preceded by the onset of the specific respiratory symptoms during the study and for whom information on the BMI before the onset of respiratory symptoms was available.
Comparisons between normally distributed continuous variables were performed using parametric tests, including t test for independent samples. Not-normally distributed variables were compared by means of the Mann-Whitney test. Comparisons between proportions were performed using the 
2 test, and the strength of the association between BMI categories and AOD was quantified by the odds ratio (OR) and tested for statistical significance using 95% confidence intervals (CIs).
Analysis 2
To study the effect of gender on the association between BMI and asthma and in order to have a sample size suitable for stratification, we compared all 169 cases of new physician-confirmed asthma, regardless of the presence of any previous AOD, to 1,903 control subjects who reported no asthma or shortness of breath with wheezing.
The association between being preobese or obese and the onset of asthma was tested in logistic regression models adjusted for other risk factors, including the presence of CB and/or emphysema. Stratification by gender and other variables was performed according to the Mantel-Haenszel method. The presence of effect modification was detected through the test for homogeneity, testing the null hypothesis that the ORs in the different strata were equal. Adjusted ORs for the association between any BMI category and asthma were also computed separately for female and male subjects. An 
= 0.05 level of significance was chosen for all the performed statistical tests both in analysis 1 and in analysis 2.
|
Results |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
28 (OR, 2.10; 95% CI, 1.31 to 3.36; and OR, 1.80; 95% CI, 1.32 to 2.46, respectively).
|
28 (as compared with BMI < 28) and asthma or CB were still significant. The adjusted ORs were even higher than the crude ORs, particularly for asthma, suggesting the presence of an inverse confounding effect.
|
28 before the onset of respiratory symptoms—the first report of "shortness of breath with wheezing" for asthma cases and of "chronic cough and phlegm" for CB cases. The respective percentages were 30.4% and 25.1% when BMI was assessed before the diagnosis of the disease. Therefore, the proportion of subjects with BMI 28 and the relative crude and adjusted ORs were similar whether the BMI levels were measured before the diagnosis of the disease or before the onset of respiratory symptoms.
Analysis 2
Table 4
shows that the magnitude of the crude and adjusted ORs for the association between BMI 28 and asthma was higher among young than older subjects, among nonatopics than atopics, among nonsmokers than smokers, and among subjects with co-presence of COPD than subjects without it. However, the test for homogeneity was not significant for any of these potential effect modifiers.
|
28 and asthma were statistically higher for women than for men, according to the test for homogeneity (p = 0.056). Being preobese or obese increased significantly the risk of acquiring asthma among female subjects (adjusted OR, 3.45; 95% CI, 2.10 to 5.67) but not among male subjects (adjusted OR, 1.69; 95% CI, 0.88 to 3.27). The effect modification by gender is also shown in Figure 1 , where the adjusted ORs for the association between any BMI category, as it refers to normal weight, and asthma are computed separately for female and male subjects. Preobese or obese females are at increased risk for acquiring asthma, whereas the ORs for preobese or obese males reach only a borderline significance.
|
|
Discussion |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
28 had an increased risk of acquiring asthma. The potential confounding effect of COPD on this association was addressed by excluding asthmatic patients with other previous AOD in analysis 1 and by adjusting the ORs for the presence of COPD in analysis 2. After excluding subjects with emphysema or CB and after adjusting for other risk factors, Young et al8 found BMI categories linearly correlate with prevalence rates of asthma in a large military population. A dose-response relationship between BMI levels and asthma was evident in other cross-sectional studies on adults,6 as well as in the prospective Nurses Health Study II.9
We found asthmatic patients more likely to be preobese or obese (BMI 28) than control subjects, but we did not find the ORs for the association with asthma linearly increase across the BMI categories. However, when we performed the stratification by gender, a BMI 28 was a significant risk factor for asthma only among female subjects, and the ORs did increase as BMI increased in the female group.
These findings are consistent with those of previous studies reporting the association between high BMI and asthma to be stronger among female than male subjects. Findings from the National Population Health Survey in Canada5
and the 1970 British Cohort Study6
showed that among adult women, but not men, the ORs for the association with asthma increased with increasing BMI categories. In the Canadian study, in the highest BMI category (BMI 28), the adjusted OR for asthma was 1.89 (95% CI, 1.33 to 2.68) for women vs 0.93 (95% CI, 0.62 to 1.38) for men. In the latter study, the OR comparing young women in the highest quintile of BMI with those in the lowest quintile was 2.00 (95% CI, 1.36 to 2.94) vs 1.39 (95% CI, 0.89 to 2.17) for men.
Our findings do not support the hypothesis that obesity is the result rather than a cause of asthma through a reduction of physical activity related to the respiratory impairment. Due to the longitudinal nature of our data, in fact, we could assess the presence of an abnormal BMI among asthmatic patients both prior to the asthma diagnosis and prior to the onset of respiratory symptoms. In both cases, the proportion of preobese or obese subjects was significantly higher than among control subjects. We identified the onset of respiratory symptoms with the first report of shortness of breath with wheezing, as these symptoms have been shown to strongly relate to asthma.11 12 However, it should be stated that the assessment of the temporal sequence between obesity and asthma is complex, as asthma can be associated to respiratory symptoms other than shortness of breath with wheezing, and it is very difficult to distinguish with any certainty incident from recrudescent asthma cases in an adult population.
Several explanations for the association between obesity and asthma are possible. First, the presence of potential confounders should be taken into account. Obese people can be more likely to be exposed to negative dietary factors, such as a deficiency of antioxidant vitamins, potentially related, in turn, to asthma.13 They could also be more likely to adopt sedentary lifestyle and to spend more time indoors. A sedentary lifestyle has been suggested to affect breathing patterns, reduce airway caliber, and increase exposure to environmental tobacco and indoor allergens strongly associated with asthma.14 15 However, obesity has been shown to be a risk factor for asthma even after adjusting for diet and physical activity.9
Alternatively, it is possible that the association between obesity and asthma is affected by misclassification. Obesity leads to an increased work of breathing and predisposes to sleep disorders. The breathlessness associated with these conditions could be misinterpreted by the patients as asthma symptoms leading to misclassification. However, many studies have used physician-confirmed asthma as main outcome or even stricter definitions, reducing the risk of misclassification.
In elderly populations, the potential misclassification of cardiac diseases as asthma cases should be also taken into account. Congestive heart failure, in fact, is one of the most common causes of dyspnea in the elderly,16 and has been associated with bronchial hyperresponsiveness.17 Obesity is a known risk factor for cardiac diseases. However, if our findings were explained by such a misclassification, we would expect the association between obesity and asthma to be stronger among the subjects at increased risk of cardiac diseases, such as in older subjects. Instead, we found the ORs for the association between obesity and asthma to be higher in the group of 20- to 49-year-old subjects than in the group of older subjects.
Several mechanisms have been proposed for a potential causal relationship between obesity and asthma. Obesity can directly affect the airway caliber through the chest wall restriction. Narrowing of airways and a reduction of lung volumes have been associated with bronchial hyperreactivity.18 Obesity also predisposes to gastroesophageal reflux disease, a known risk factor for asthma.
According to these potential mechanisms, however, the effect of obesity on asthma would be expected to be similar in both sexes, whereas the association between obesity and asthma was significant only among women. The effect modification by gender suggests that sex hormones could be involved in the causal pathway. Estrogen and progesterone levels are affected by obesity and are related to asthma as well. By assessing BMI at age 11 years and presence of wheezing at age 11 years and 13 years, Castro-Rodriguez et al2 found that the association between overweight status and asthma was stronger among girls whose puberty started before the age of 11 years than among girls whose puberty started later. Also, among adult females, at least postmenopausal women, BMI has been shown positively associated with plasma estrogen that, in turn, could increase the risk of acquiring asthma.19
We found emphysema associated with a BMI < 18.5 and CB associated with a BMI 28. The temporal sequence between abnormal BMI and COPD (emphysema or CB) is complex to assess. COPD and smoking, in fact, are likely to affect weight according to a bidirectional relationship. We assessed BMI before the diagnosis of COPD and, for patients with CB, before the onset of chronic cough and phlegm (clinical definition20
). However, as the onset of smoking habits dated back to the years preceding the enrollment in the study for most COPD patients, we could not assess the temporal sequence of smoking, abnormal BMI, and onset of COPD, despite the longitudinal nature of our data. Further research is required to clarify this complex interrelationship.
The association between emphysema and malnutrition is known but not fully understood. Recent findings from the Baltimore Longitudinal Study of Aging21 suggest that low BMI could itself be a risk factor for COPD. In this study, men in the lowest BMI tertile at baseline were almost three times more likely to receive a diagnosis of COPD during the follow-up period than men in the highest tertile. In another study,22 underweight patients with COPD were found to have lower carbon monoxide diffusing capacity and higher dyspnea scores than normal-weight patients with COPD. These characteristics are traditionally attributed primarily to the more emphysematous form of COPD.23 Emphysema patients can progressively lose weight due to several mechanisms. For instance, it has been hypothesized that the oxygen cost of breathing is increased in these patients,24 25 and the caloric intake is reduced since large meals can induce shortness of breath. Although we assessed the BMI before the AOD diagnosis in our study, it is likely that most of the patients with emphysema experienced a gradual onset of the pulmonary obstructive disease years before receiving the diagnosis. In this case, the effect of the disease on the nutrition status, rather than the effect of malnutrition on the disease, could explain the association between low BMI and emphysema.
Our data also confirm that the association between BMI and COPD is largely affected by the type of diagnosis. In a prospective epidemiologic study, Higgins et al26
found leanness to have a predictive value for developing AODs among men but not women. Similarly, in a cross-sectional study,1
a BMI < 20 among male subjects and a BMI 28 among female subjects were associated with an increased prevalence of COPD. However, since no distinction was made between emphysema and CB, it is possible that these findings were related to the different distribution of the two types of COPD within the genders. Male subjects, in fact, have been shown to be more likely to have a diagnosis of emphysema and female subjects a diagnosis of CB.27
In our study, we broke down the COPD group according to the two diagnoses (emphysema or CB) and found emphysema associated with low BMI categories and CB with high BMI categories. However, we acknowledge that an epidemiologic distinction between the two diseases is difficult and that a physician’s diagnosis of COPD itself could be influenced by the weight of the patients, according to the traditional distinction of "plethoric" CB and "thin" emphysema. The risk of misclassification may be particularly relevant for CB patients. We have previously shown that the vast majority of subjects with new onset of CB do not satisfy the symptom criteria of chronic cough and phlegm at the first report of the disease,28
and, consistently, less than one third of our CB cases had chronic cough and phlegm at any time before receiving the diagnosis. Nevertheless, by restricting the analysis to these patients with CB, we were still able to confirm the association between obesity and CB. Of note, by combining patients with CB and patients with emphysema in a single COPD category, we would find a U-shaped risk trend, meaning that both low and high BMIs increase the risk for the disease.
To reduce the risk of misclassification, we selected only physician-confirmed cases, but the physician-confirmed diagnosis was still reported by the patient and could be sensitive to recall bias and lack of specificity or sensitivity. This problem cannot be completely ruled out, as some limitations to the specificity and sensitivity of case definition are intrinsic to the design of large epidemiologic studies.
In conclusion, emphysema is associated with being underweight; asthma and CB are associated with being preobese or obese. The temporal sequence between smoking, abnormal BMI, and development of COPD (emphysema or CB) is complex to assess. In contrast, the development of asthma seems to be temporally associated to a BMI 28, as patients with asthma are more likely to be obese both after and before the onset of respiratory symptoms. The association between elevated BMI and asthma is significant only among female subjects.
|
Acknowledgements |
|---|
|
Footnotes |
|---|
At the time of the study, Dr. Guerra was a fellow from the Institute of Respiratory Diseases, IRCCS Policlinico Hospital, University of Milan, Italy.
Drs. Guerra, Sherrill, and Martinez were funded by grant No. HL 56177 from the National Heart, Lung, and Blood Institute.
Received for publication November 5, 2001. Accepted for publication April 16, 2002.
|
References |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. Tomoda, M. Yoshikawa, T. Itoh, S. Tamaki, A. Fukuoka, K. Komeda, and H. Kimura Elevated Circulating Plasma Adiponectin in Underweight Patients With COPD Chest, July 1, 2007; 132(1): 135 - 140. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Chinn, S. H. Downs, J. M. Anto, M. W. Gerbase, B. Leynaert, R. de Marco, C. Janson, D. Jarvis, N. Kunzli, J. Sunyer, et al. Incidence of asthma and net change in symptoms in relation to changes in obesity Eur. Respir. J., October 1, 2006; 28(4): 763 - 771. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Chen, R. Dales, and Y. Jiang The association between obesity and asthma is stronger in nonallergic than allergic adults. Chest, September 1, 2006; 130(3): 890 - 895. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Kostikas, A. I. Papaioannou, K. I. Gourgoulianis, C. J. Etzel, S. Kachroo, H. Yanai, H. Yoshida, N. Tada, B. C. Jacobson, and C. A. Camargo Jr. BMI and Gastroesophageal Reflux in Women N. Engl. J. Med., August 24, 2006; 355(8): 848 - 850. [Full Text] [PDF]
|
|
|
|
|
|
L. Watson, J. P. Schouten, C-G. Lofdahl, N. B. Pride, L. A. Laitinen, D. S. Postma, and on behalf of the European Respiratory Society Stud Predictors of COPD symptoms: does the sex of the patient matter? Eur. Respir. J., August 1, 2006; 28(2): 311 - 318. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Poulain, M. Doucet, G. C. Major, V. Drapeau, F. Series, L.-P. Boulet, A. Tremblay, and F. Maltais The effect of obesity on chronic respiratory diseases: pathophysiology and therapeutic strategies. Can. Med. Assoc. J., April 25, 2006; 174(9): 1293 - 1299. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A Sood, E S Ford, and C A Camargo Jr Association between leptin and asthma in adults Thorax, April 1, 2006; 61(4): 300 - 305. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Peters-Golden, A. Swern, S. S. Bird, C. M. Hustad, E. Grant, and J. M. Edelman Influence of body mass index on the response to asthma controller agents. Eur. Respir. J., March 1, 2006; 27(3): 495 - 503. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S Deane and A Thomson Obesity and the pulmonologist Arch. Dis. Child., February 1, 2006; 91(2): 188 - 191. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. A. Johnston, T. A. Theman, and S. A. Shore Augmented responses to ozone in obese carboxypeptidase E-deficient mice Am J Physiol Regulatory Integrative Comp Physiol, January 1, 2006; 290(1): R126 - R133. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Iversen, P. C. Hannaford, D. B. Price, and D. J. Godden Is Living in a Rural Area Good for Your Respiratory Health?: Results From a Cross-sectional Study in Scotland Chest, October 1, 2005; 128(4): 2059 - 2067. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Chen, D. Rennie, Y. Cormier, and J. Dosman Sex Specificity of Asthma Associated With Objectively Measured Body Mass Index and Waist Circumference: The Humboldt Study Chest, October 1, 2005; 128(4): 3048 - 3054. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A Johannessen, E R Omenaas, P S Bakke, and A Gulsvik Implications of reversibility testing on prevalence and risk factors for chronic obstructive pulmonary disease: a community study Thorax, October 1, 2005; 60(10): 842 - 847. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. J Vieira, A. M Ronan, M. R Windt, and A. R Tagliaferro Elevated atopy in healthy obese women Am. J. Clinical Nutrition, September 1, 2005; 82(3): 504 - 509. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. G. Sulit, A. Storfer-Isser, C. L. Rosen, H. L. Kirchner, and S. Redline Associations of Obesity, Sleep-disordered Breathing, and Wheezing in Children Am. J. Respir. Crit. Care Med., March 15, 2005; 171(6): 659 - 664. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Iribarren, I. V Tolstykh, and M. D Eisner Are patients with asthma at increased risk of coronary heart disease? Int. J. Epidemiol., August 1, 2004; 33(4): 743 - 748. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Guerra, A. L. Wright, W. J. Morgan, D. L. Sherrill, C. J. Holberg, and F. D. Martinez Persistence of Asthma Symptoms during Adolescence: Role of Obesity and Age at the Onset of Puberty Am. J. Respir. Crit. Care Med., July 1, 2004; 170(1): 78 - 85. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P A Dawkins, L J Dowson, P J Guest, and R A Stockley Predictors of mortality in {alpha}1-antitrypsin deficiency Thorax, December 1, 2003; 58(12): 1020 - 1026. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
