A REGRESSION APPROACH TO THE ANALYSIS OF SERIAL PEAK FLOW AMONG FUEL-OIL ASH EXPOSED WORKERS

We investigated the association between exposure to fuel oil ash and a cute airway obstruction in 31 boilermakers and 31 utility workers duri ng the overhaul of a large oil-fired boiler. Air flow was assessed wit h self-recorded serial peak expiratory flow rate measurements (PEFR) u sing a mini-Wright meter. Exposure to thoracic particulates with an ae rodynamic diameter of 10 mu m or smaller (PM(10)) was assessed using p ersonal sampling devices and detailed work diaries. All subjects were male, with an average age of 43 yr, and an average of 18 yr at their c urrent trade. Average PM(10) exposure on work days was 2.75 mg/m(3) fo r boilermakers and 0.57 mg/m(3) for utility workers. Three daily PEFR measurements (start-of-shift, end-of-shift, and bed-time) were analyze d simultaneously, using Huber linear regression. After adjustment for job title, welder status, age, height, smoking, and weld-years, for ea ch mg/m(3) increase in PM(10), the estimated decline in PEFR was 13.2 L/min (p = 0.008) for end-of-shift, 9.9 L/min (p = 0.045) for bed-time , and 6.6 L/min (p = 0.26) for start-of-shift of the following day. Th is decline of the exposure effect over the 24-h period that follows wa s statistically significant (p = 0.004). No other factors were found t o significantly modify the effect of exposure. Our results suggest tha t occupational exposure to fuel oil ash is associated with significant acute decrements in peak flow.