Lung Function before and after a Large Chlorine Gas Release in Graniteville, South Carolina

Abstract

Rationale: On January 6, 2005 a train derailment led to an estimated 54,915-kg release of chlorine at a local textile mill in Graniteville, South Carolina.

Objectives: We used the employee health spirometry records of the textile to identify enduring effects of chlorine gas exposure resulting from the incident on the lung function of workers employed at the textile mill.

Methods: Spirometry records from 1,807 mill workers (7,332 observations) were used from 4 years before and 18 months after the disaster. Longitudinal analysis using marginal regression models produced annual population mean estimates for FEV1, FVC, and FEV1/FVC ratio. Covariate adjustment was made for sex, age, smoking, height, season tested, technician, obesity, season × year interactions, and smoker × year interactions. The increased prevalence of mill workers having accelerated FEV1 decline was also evaluated after the chlorine spill.

Measurements and main results: In the year of the accident, we observed a significant reduction in mean FEV1 (-4.2% predicted; P = 0.019) when compared with the year before the incident. In the second year, partial recovery in the mean FVC % predicted level was seen, but the cohort's average FEV1/FVC ratio continued to decrease over time. Severe annual FEV1 decline was most prevalent in the year of the accident, and independent of mill worker smoking status.

Conclusions: The Graniteville mill worker cohort revealed significant reductions in lung function immediately after the chlorine incident. Improvement was seen in the second year; but the proportion of mill workers experiencing accelerated FEV1 annual decline significantly increased in the 18 months after the chlorine incident.

Keywords: Graniteville; chlorine gas; disaster epidemiology; lung function; rapid FEV1 decline; toxic gas exposure.

Figure 1.

Selection of the Graniteville mill-worker cohort analytical data set, based on the study inclusion criteria. Obs = observations; PFT = pulmonary function testing.

Figure 2.

Annual trends in mill worker spirometry (i.e., means with 95% confidence interval) for FVC volume, FEV₁ volume, and FEV₁/FVC ratio. Covariate adjustment was made for sex, height, race, obesity, smoking, testing technician, season of testing, and season × year effect modification. Adapted by permission from Reference .