An epidemiological investigation of the role of phenotype in the association of obesity and asthma.

Abstract

This thesis investigates the complexity in the relationship between obesity and asthma and asthma morbidity. Previous epidemiological studies exploring these relationships have been limited by sample bias and the use of restricted phenotypes of body mass index (BMI) and self-reported asthma, ignoring the problem of undiagnosed asthma, and more pathogenic central obesity phenotypes. Cardiovascular disease (CVD), a systemic manifestation of obesity may be augmented by asthma-related airway inflammation, yet studies inconsistently identifying an association with asthma have failed to assess the role of asthma phenotype or cardiotoxic effects of short acting beta-2 adrenergic agonists (SABA). Understanding the consequences of this complexity is fundamental to the development of appropriate policy and intervention. The North West Adelaide Health Study, a representative biomedical population sample (n=4006) permitted an examination of the role of phenotype in the association of obesity [body mass index (BMI), waist circumference, waist to hip ratio] with asthma [atopy, significant bronchodilator reversibility (SBR)]. Optimising the identification of asthma in the absence of a gold standard test is important. The prevalence of undiagnosed asthma (SBR in absence of doctor diagnosis) was variable (1.6% to 4.5%) depending on the SBR criteria specified. The observed symptom burden and lung function impairments suggest that all criteria identified subjects with probable asthma. SBR criteria were associated with different sociodemographic factors and the 9% of the predicted criterion was least biased particularly in terms of age and sex. Generalised (BMI) and central obesity were associated with asthma in females only. After consideration of atopic status, in males, central obesity and high BMI (likely to be distributed centrally) was associated with non-atopic asthma. In females central obesity was also associated with non-atopic asthma but a high BMI was associated with atopic asthma. This suggests different pathophysiological mechanisms for the relationship between obesity and atopic and non-atopic asthma. In subjects with asthma, a significant burden of generalised and central obesity-related asthma morbidity (symptoms, beta-2 agonist use, lung function) occurred largely in males only, although quality of life impairments and increased primary care visits were not sex-specific. Only central obesity was associated with persistent airways obstruction in males. Asthma was associated with CVD/stroke events, independent of traditional CVD risk factors in cross-sectional analyses. Asthma was not associated with diabetes or cardiovascular risk factors. No modifying effect of obesity was observed in these associations, suggesting that events may be related to aspects of asthma pathology, asthma phenotype or a direct cardiotoxic effect of SABA. In females, incident CVD/stroke events were associated with asthma and as required SABA use, but the association was not modified by atopic status. In males, CVD/stroke events were associated with other respiratory morbidity. Few events occurred in men with asthma, but a significant interaction of asthma with atopic status was evident. This work has contributed to emerging knowledge that improved phenotyping will advance our understanding of the relationship and mechanisms between obesity and asthma and has implications for asthma management. An unbiased SBR criterion will improve the identification of asthma in the absence of a gold standard test. The association of central obesity with non-atopic asthma indicates that asthma should be considered in such symptomatic individuals. Given the increased morbidity burden in obese subjects with asthma, healthy weight maintenance is an important component of asthma management. Management of macrovascular disease risk in women with asthma includes caution in the prescribing of SABA.