Independent Clinical and Occupational Predictors of Moderate-to-Severe Occupational Asthma: A Retrospective Cross-Sectional Study from Iraq

Abstract

Occupational asthma (OA) remains one of the most frequent work-related respiratory disorders, yet the determinants of greater disease severity are poorly characterised, especially in low- and middle-income settings such as Iraq. This retrospective cross-sectional study, conducted at Baghdad Teaching Hospital between August and October 2025, examined 158 adults with a clinical diagnosis of OA to identify clinical and occupational predictors of moderate-to-severe disease. Severity was classified using spirometric indices (FEV1 and FEV1/FVC ratio) as mild or moderate-to-severe, without incorporating symptom control, exacerbation frequency, or treatment intensity. Disease was mild in 52.5% of patients, moderate in 41.8%, and severe in 5.7%, so nearly half (47.5%) of the cohort fell into the moderate-to-severe category. Bivariate comparisons showed that ever smoking, occupational rhinitis, and a longer duration of symptoms before diagnosis were significantly more common among moderate-to-severe cases. Multivariable logistic regression confirmed these three variables as independent predictors: ever smoking (adjusted odds ratio 2.91), occupational rhinitis (adjusted odds ratio 2.35), and each additional year of symptom duration (adjusted odds ratio 1.15). Age, sex, body mass index, atopic status, and the class of causative occupational agent showed no significant association with severity. The regression model showed good calibration (Hosmer-Lemeshow P = 0.667) and acceptable apparent discrimination (AUC 0.812), though this was not externally validated. These findings suggest that smoking behaviour, coexisting occupational rhinitis, and delayed diagnosis are the principal modifiable contributors to OA severity, underscoring the need for earlier workplace screening, smoking-cessation interventions, and proactive management of upper-airway disease among exposed workers in Iraq.