Differential association between educational levels and machine learning-derived asthma phenotypes in adults

<bold>Background:</bold> The influence of socioeconomic gradients on asthma has been recognized for decades, but socioeconomic status (SES) variations in asthma different asthma phenotypes remains underexplored. <bold>Aim:</bold> To examine SES variations, as determined by educational attainment and occupational skill levels, in machine learning-derived asthma phenotypes in adults. <bold>Methods:</bold> From the West Sweden Asthma Study, machine learning classified 1,895 asthmatics, aged 16-75, into four phenotypes: burdensome late-onset non-atopic asthma with smoking, female-dominated early adult-onset asthma, adult-onset asthma with high inflammation, and early-onset mild atopic asthma. we investigated socioeconomic status' effect (education and occupational class) on these phenotypes using adjusted mutinominal regression, referencing 1,206 non-asthmatics. <bold>Results:</bold> Compared to primary education, secondary education was associated with higher odds of severe late-onset non-atopic asthma with smoking (OR 1.61, 95%CI 1.26-2.07). Secondary (OR 0.44, 95%CI 0.29-0.66) and tertiary education (OR 0.62, 95%CI 0.47-0.82) were associated with a decreased likelihood of the female-dominated early adult-onset asthma phenotype compared to primary education. Tertiary compared to primary education was also associated with lower odds of early-onset mild asthma phenotype. Occupational skill levels were not associated with any of the asthma phenotypes. <bold>Conclusion:</bold> Education, rather than occupation, as an indicator of SES, was differentially associated with asthma phenotypes in adults.