Objectives In this longitudinal study we aimed to estimate the longitudinal fall in forced expiratory volume at one second (FEV₁) associated with different levels of thoracic aerosol exposure in cement production workers.

Methods 4998 workers in 24 plants in eight countries were included in the study. Personal measurements of thoracic dust (n = 6942) were performed in 24 plants for 8 job types in 2007, 2009 and 2011. Arithmetic mean (AM) exposure in job types and plants was estimated from mixed models taking year and seasonal variation into account. FEV₁ was measured by spirometry in the same years. The longitudinal change in FEV₁/m² of standing height was analysed in mixed models allowing for age, sex, smoking, allergy, asthma, previous dust exposure, and use of airway protection, using estimated exposures multiplied by observational time.

Results The AM exposure level among non-administration employees varied between job types and plants from 0.14 to 14 mg/m³. A longitudinal loss in FEV₁/m² of 2.9 ml/m²*(yr*mg/m³) (95% Confidence Interval, CI, 2.5–3.2) was estimated, assuming linear relationships. Using exposure stratified in quintiles, losses per (mg/m³)*yr were 5.9 ml (CI 3.0–8.8) in the 0.14–0.61 mg/m³ category (reference), 9.7ml (CI 7.6–12) in the 0.62–1.57 mg/m³ category, 14 ml (CI 12–17) in the 1.58–2.27 mg/m³ category, 18 ml (CI 15–20) in the 2.28–5.40 mg/m³ category and 19 ml (CI 15–22) in the 5.41–14 mg/m³ category.

Conclusions FEV₁ loss increased with exposure indicating that cement aerosol exposure may lead to adverse lung changes. Interpreting the change in the lowest exposure category as the age effect, FEV₁ losses in all other categories are significantly greater. Assuming that the thoracic fraction is 45–50% of total dust, the second lowest exposure level (0.62–1.57 mg/m³) is well below the present occupational exposure limit of 10 mg/m³ of total dust.