A survey of actinic keratoses among paraquat production workers and a nonexposed friend reference group

Follow-up on two reports of an excess of keratoses among paraquat production workers was conducted to evaluate the contribution of occupational exposures to the prevalence of keratoses among workers in a paraquat production plant in Texas. A cross-sectional study design was used to compare the prevalence of keratoses among current workers to an age, race, and sex frequency-matched group of their friends who had never worked at the plant. The analysis is based on 112 workers and 232 friends. Exposure, outcome, and covariables used in the analyses were obtained from an interview questionnaire, dermatology exam, and company records. Overall, the prevalence proportion of the presence of any actinic keratoses among workers and friends was similar (0.30 and 0.28, respectively). Among high cumulative exposed workers, the prevalence of any actinic keratoses was 0.40 compared to 0.20 among low-exposed workers and 0.28 among friends. These results were further explored using a multiple logistic regression approach to adjust for known risk and possibly confounding variables. Statistically significant high risks of actinic keratoses were demonstrated for freckling before age 16 years, older age, suntanning behaviors, occupational exposure to polycyclic aromatic hydrocarbons (PAH), and Fitzpatrick skin type. There was no significant contribution of overall exposure status (worker vs. friend). Similar to the crude analysis, the odds of actinic keratoses of high-exposed workers compared to friends was 1.9 (confidence interval [CI] = 0.9-4.2) whereas the comparable odds ratio for low-exposed workers vs. friends was 0.6 (CI = 0.2-1.7). These data do not demonstrate an excess of actinic keratoses overall nor any consistent increase in the odds of keratoses with an increase in plant exposure level.