Associations Between Occupational Exposures and Cough Subclasses Among Middle-Aged Australians.

Objective: The evidence around occupation-related chronic cough is conflicting and current definitions of chronic cough cannot capture its heterogeneity. Using our recently characterised novel cough subclasses, we aimed to identify subclass-specific occupational risks.

Methods: Using data from the Tasmanian Longitudinal Health Study (TAHS), occupational exposures up to age 53 years were coded using the ALOHA+ Job Exposure Matrix, into ever-exposure (no, only-low, ever-high) and cumulative exposure. People belonging to six previously identified cough subclasses among 2213 current coughers at age 53 years were compared to non-coughers (n = 1396). Associations with occupational exposures were assessed using multinomial logistic regression for these cough subclasses and logistic regression for standard definitions (chronic cough, chronic phlegm, and chronic bronchitis) after adjusting for potential confounders.

Results: Biological dust was associated with "cough with allergies" (cumulative: adjusted multinomial odds ratio [aMOR] = 1.06, 95% CI: 1.02-1.10, per 10 exposure-year increase). Aromatic solvents were associated with "chronic dry cough" (cumulative: aMOR = 1.15, 95% CI: 1.02-1.29). Other solvents were associated with "chronic productive cough" (ever-high: aMOR = 2.81, 95% CI: 1.26-6.2); "intermittent productive cough" (cumulative: aMOR = 1.06, 95% CI: 0.98-1.16), chronic bronchitis (ever-high: aOR = 2.48, 95% CI: 1.01-6.06); and chronic phlegm (ever-high: aOR = 2.26, 95% CI: 1.14-4.51). Herbicides (cumulative) were also associated with "intermittent productive cough" (aOR = 1.09, 95% CI: 1.00-1.77) and chronic phlegm (aOR = 1.07, 95% CI: 1.00-1.15).

Conclusions: Novel cough subclasses had distinct associations with specific occupational exposures, suggesting different pathophysiology. Aromatic solvents were associated with dry cough; biological dust with allergic cough; herbicides and other solvents with productive cough. Using novel cough subclasses was superior to standard definitions in uncovering these associations.