Phenotypes of Airway Diseases in Adults and Variation by Socioeconomic Status

Abstract

Background: Lately, global prevalence of chronic obstructive airway diseases has risen, though some regions have seen stability. These diseases are heterogenous, have diverse manifestations and varying phenotypes. Ongoing research aims to identify and characterize phenotypes of obstructive airway diseases to enhance understanding of risk factors and improve treatments. The breadth and diversity of currently uncovered phenotypes of obstructive airway diseases are unclear. In the Nordic countries, there are social disparities in airway diseases, with those at lower socioeconomic status facing higher morbidity and mortality. Various socioeconomic indicators have been used to link socioeconomic status (SES) to airway diseases, but their impact on distinct asthma phenotypes remains unclear. Smoking is a risk factor for COPD, but its role in asthma remains controversial. While much is known about the independent impact of SES and smoking on obstructive airway diseases, their mutual synergism is unclear, particularly in relation to phenotypes of obstructive airway diseases. The goal of this thesis was to summarize the phenotypes of obstructive airway diseases and explore the connections between SES and phenotypes of airway diseases. Objectives: The overarching aim of the thesis was to characterize phenotypes of obstructive airway diseases and determine their variation by SES. The thesis consisted of 6 papers: (1) a study of association of level of education and asthma control in adult-onset asthma; (2) a study on the effect modification between smoking and socioeconomic status towards risk of obstructive airway diseases; and (3) a systematic review of studies on computational phenotyping of airway disease using unsupervised computational methods; (4) a study on association between social status and forms/phenotypes of rhinitis; (5) a study on phenotyping asthma in general population using machine learning approach and (6) a study on association between socioeconomic status and phenotypes of asthma. Methods: In Paper 1, individuals with adult-onset asthma were examined in three studies: the Obstructive Lung Disease In Northern Sweden (OLIN) (n = 593), Seinäjoki Adult Asthma Study (SAAS) (n = 200), and West Sweden Asthma Study (WSAS) (n = 301) during 2009-2014. Education levels were primary, secondary, or tertiary, with uncontrolled asthma defined as asthma control test (ACT) score ≤19. Eight hundreds ninety-six participants with complete data were included (OLIN n = 511, SAAS n = 200, WSAS n = 185). Paper 2 analyzed data from WSAS (23,753 participants) and OLIN (6,519 participants) aged 20-75. Bayesian analysis explored smoking and socioeconomic status effect modification on respiratory outcomes. Paper 3 was a systematic review that synthesized studies on computational phenotyping of airway disease published 2010-2020. In Paper 4 and 6, estimated the association between education and occupational classes and risk of forms of rhinitis and asthma phenotypes, respectively. Paper 5 employed a deep clustering algorithm for clustering asthma, that was defined based on self-report from general population. Results: In Paper 1, the odds ratio (OR) for primary education in relation to uncontrolled asthma was 1.92 (95% CI 1.15-3.20) in daily ICS users and 3.42 (95% CI 1.30-8.96) in non-atopic subjects. In Paper 2, SES and smoking had marginal synergistic effect on respiratory outcomes, with more pronounced effect-modification seen among lower occupational classes regarding smoking effect on allergic asthma, while among professionals, it affected non-allergic asthma. In Paper 3, there were variations in the literature regarding computational phenotyping of obstructive airway diseases, particularly in study design and methods, study settings, participant profiles, and variables used to perform the phenotyping exercise, altogether leading to differences in characterized phenotypes. In paper 4, higher education and occupation levels were identified to be risk factors for allergic and chronic rhinitis. Paper 5 identified four asthma phenotypes that were distinguishable by age of onset, severity, risk factors, and prognosis. Paper 6 showed an association between high and low education levels and presenting certain asthma phenotypes. Conclusion: In affluent Nordic countries, classical phenotypes of asthma are present in the general population. Lower education poses a risk for uncontrolled adult-onset asthma, while high education levels were linked to high risk of allergic and chronic forms of rhinitis, suggesting the significance of incorporating social aspect into management and prevention of airway diseases.