Subjective symptoms due to solvent mixtures, dioxin, and toluene: impact of exposure versus personality factors.

In this study, we analyse the impact of personality factors on the frequency of self-reported symptoms for workers under different exposure conditions. Reported symptoms may depend on the level and type of exposure, as well as on personality factors such as trait anxiety of the worker or his general sensitivity with regard to the environment. The employed data stems from three studies: The first study contains information of 60 workers who suspected to be exposed to polychlorined dibenzodioxins and dibenzofuranes (Lifetime Weighted Average Exposure, LWAE, as an index for contact with the substances). The second study concerns 40 workers who are exposed to different concentrations of solvent mixtures in paint manufacturing (LWAE of total hydrocarbons about 10 ppm). The third study includes repeated measurements of two subgroups of workers from rotogravure printing plants who are exposed to different concentrations of toluene: a "high" exposure group (n = 129, LWAE about 46 ppm, current exposure 25 ppm) and a "low" exposure group (n = 96, LWAE for toluene about 9 ppm, current exposure 3 ppm). Trait anxiety, environmental sensitivity, and self-reported symptoms are measured by validated questionnaires and age as well as verbal intelligence are controlled. To determine the effect of the individual characteristics and the different exposures on self-reported symptoms, frequency analyses and variance analyses are conducted and linear models are fitted. For all analyses, trait anxiety explains the highest share of the variance. If there is no effect of the exposure on the reported symptoms (dioxin and low-level toluene study), trait anxiety seems to have a larger explanatory power in comparison with those studies where the exposure has an effect on the reported symptoms (solvent-mixture and high-level toluene study). Neurotoxicological risk analysis has to account for the detected dependence of self-reported symptoms on personality traits: assessments for elevated symptoms should not only be linked to the intensity of exposure but also related to benchmarks derived from the normal variability of personality factors.