This meta-analysis has provided a precise estimate of the odds of incident asthma for individuals who are overweight or obese, suggesting that the odds of incident asthma are increased 50% in overweight or obese individuals as a whole. Our findings also demonstrate a clear dose–response relationship between BMI and asthma, suggesting that asthma risk increases further as body weight increases. In addition, we have shown that female sex does not seem to disproportionately affect the obesity–asthma relationship, given that the odds of incident asthma in overweight and obese men and women were similar.
On the basis of these findings, overweight and obesity seem to be significant risk factors for asthma, and if they can be considered to be modifiable risk factors, interventions that effect weight loss could be associated with a decrease in asthma incidence. Survey data suggest that two-thirds of the adult population of 220 million in the United States are overweight or obese (2
). Assuming that approximately 12% of these overweight or obese individuals have asthma (25
) and assuming that 6% of the remaining adult population has asthma, approximately 200 million United States adults are free of asthma but at risk of developing it. Studies suggest that new asthma cases in the general adult population occur at a rate of approximately 0.5% per year (27
), and presumably this rate is influenced by contributions from the lean and overweight/obese subgroups of the population in the ratio described previously (1 lean:2 obese/overweight). If significant weight loss could be achieved in the population of overweight and obese individuals, it could be estimated that the number of new asthma cases in United States adults might fall by as much as 250,000 per year (from 0.5 to 0.375% per year). If these rates of increase can be extrapolated to the pediatric population, where the annual incidence of asthma is as much as five times higher (24 per 1,000 person-years ), the effect of even small changes in mean population BMI may translate into significant increases or decreases in asthma incidence in children and adults.
This analysis has a number of potential limitations. First, many of the studies included in this meta-analysis relied on self-reported, rather than measured, weight and height to determine BMI. If there were large and systematic differences in reporting of weight and height by sex or BMI, then these results could be confounded by a classification bias. We believe this is unlikely to have affected our results. Of the seven studies included in this analysis, two (11
) used measured and not self-reported height and weight, and these studies found that the relationship between obesity and asthma was significant and was similar between men and women, which is consistent with our results. In addition, a recent study by Hu and colleagues validated self-reported with measured weight among 184 participants in the Nurses' Health Study (30
). Self-reported weight was highly correlated with measured weight (r
= 0.96; mean measured weight was 1.4 kg greater than self-reported weight). This analysis was performed in a study cohort included in this meta-analysis (10
BMI may not be the best measure of adiposity, particularly when looking at the effect of obesity on lung disease. There may be sex differences in muscle mass and body fat distribution that could make BMI a misleading indicator of the degree of adiposity, and measures of abdominal adiposity are better predictors of alteration in pulmonary function than body weight or BMI (31
). Although future research in obesity and lung disease may benefit from the use of alternative measures of adiposity, for practical purposes we were constrained to use BMI because it is by far the most commonly used measure of obesity in this literature.
It is possible that asthma may be overdiagnosed in an obese population or that the phenotype of asthma seen with overweight and obesity is unique regarding clinically meaningful parameters, such as the nature or perception of symptoms, specific physiologic impairments (33
), or response to therapy (34
). Although the data on which our analysis is based do not allow evaluation of these factors in the combined study population, previous reports suggest that asthma associated with obesity may differ phenotypically from asthma in normal-weight individuals. Obesity in the absence of asthma causes physiologic impairments in lung function, including reduction in lung volumes (35
), chest wall restriction (36
), and increased oxygen cost of breathing (37
), and contributes to comorbid conditions such as gastroesophageal reflux (38
) and sleep apnea; these can result in dyspnea and wheezing, which might be mistaken for asthma by patients and clinicians (33
). It has also been reported that although lung volumes are reduced and asthma symptoms are increased in obesity, airflow obstruction and airway hyperresponsiveness are not altered (41
). Weight loss studies have shown improvements in lung function and asthma symptoms (42
) but not necessarily in airflow obstruction or airway hyperresponsiveness (43
). These reports cast some doubt on the validity of self-reported asthma (even if it is also “physician diagnosed”) in large epidemiologic studies. It is reasonable to believe that some of these patients with “asthma” may have respiratory symptoms due to obesity but may not meet rigorous objective physiologic criteria for asthma (40
). This type of classification bias is difficult to address in a large epidemiologic investigation without having independent clinical and physiologic data for each subject and may have falsely inflated the number of new cases of asthma in obese subjects, leading to an overestimate of the OR, a phenomenon that might contribute to the overdiagnosis of asthma in reports in which conclusions are drawn from epidemiologic data using self-reported asthma as the primary criterion for diagnosing asthma, without supporting physiologic evaluation.
Alternatively, the calculated ORs may have been underestimated due to the grouping together of underweight and normal-weight subjects. Being underweight has been shown to be a risk factor for asthma, with the relationship between asthma risk and BMI having a J-shape such that very low BMI has been reported to be associated with elevated asthma risk (15
). By defining “normal” BMI as less than 25, we potentially included underweight individuals with a high incidence of asthma in the “normal” group, which elevates the group's asthma incidence and diminishes the relative odds of asthma due to obesity. Because three of the seven studies included BMI of less than 20 in their “normal” group, we were unable to rigorously exclude underweight individuals in our analysis.
Although the effect of obesity on asthma in this study was statistically significant and there was a clear dose–response relationship, the magnitude of the effect was modest, even if one assumes that all newly reported cases of asthma in these studies are asthma and not obesity-associated respiratory symptoms. However, to standardize data for our analysis, we determined the number of new cases of asthma per year; thus, the increased odds are elevated even over a relatively short time frame with regard to duration of overweight and/or obesity in many patients and when multiplied over many years are likely to be clinically significant.
Obesity is a well-established risk factor for diabetes, hypertension, sleep apnea, stroke, cardiovascular disease, arthritis, cancer, and many other diseases (44
). Our findings support adding asthma to this list and should provide yet one more piece of information to compel obese individuals to lose weight and to support the aggressive implementation of public health measures to support the attainment of this goal.