Head-space gas chromatography (GC) and high-performance liquid chromatography (HPLC) (with fluorescence detectors) methods were developed for toluene (TOL-U) and o-cresol (CR-U) in urine, respectively. In order to identify the most sensitive urinary indicator of occupational exposure to toluene vapor (TOL-A) among TOL-U, CR-U, and hippuric acid in urine (HA-U), the two methods together with an HPLC (with untraviolet detectors) method for determination of HA-U were applied in the analysis of end-of-shift urine samples from 115 solvent-exposed workers (exposed to toluene at 4 ppm as geometric mean). Regression analysis showed that TOL-U correlated with TOL-A with a significantly higher correlation coefficient than did HA-U or CR-U. With regard to the TOL-A concentrations at which the exposed subjects could be separated from the nonexposed by the analyte, TOL-U achieved separation at < 10 ppm TOL-A, whereas both HA-U and CR-U did so only when TOL-A was 30 ppm or even higher. The ratio of the analyte concentrations at 50 ppm TOL-A to those at 0 ppm TOL-A was also highest for TOL-U. Overall, the results suggest that TOL-U is a better marker of exposure to toluene vapor than HA-U or CR-U.

One hundred and forty-three workers exposed to one or more of toluene, xylene, ethylbenzene, styrene, n-hexane, and methanol at sub-occupational exposure limits were examined for the time-weighted average intensity of exposure by diffusive sampling, and for biological exposure indicators by means of analysis of shift-end blood for the solvent and analysis of shift-end urine for the corresponding metabolite(s). Urinalysis was also performed in 20 nonexposed control men to establish the "background level." Both solvent concentrations in blood and metabolite concentrations in urine correlated significantly with solvent concentrations in air. Comparison of blood analysis and urinalysis as regards sensitivity in identifying low solvent exposure showed that blood analysis is generally superior to urinalysis. It was also noted that estimation of exposure intensity on an individual basis is scarcely possible even with blood analysis. Solvent concentration in whole blood was the same as that in serum in the case of the aromatics, except for styrene. It was higher in blood than in serum in the case of n-hexane, and lower in the cases of styrene and methanol.

SUMMARY To clarify the relationship between morphological changes in the corneal endothelium and ultraviolet (UV) radiation, specular microscopic examinations were performed on both eyes of 118 welders and 85 controls. The results showed: a decrease in the hexagonal cells in welders (20-29 years) in comparison with the controls (20-29) (p<005); an increase in the mean cell size of the endothelium and a decrease in the hexagonal cell population with increasing age in both groups; increases in standard deviation (SD) and the coefficient of variation (CV) of the mean cell size in both groups; increases in SD and CV of the mean number of cell sides in both groups; and no difference in the mean cell size between the two groups. These results show that UV radiation damages not only the corneal epithelium but also the endothelium, and suggest that it causes more pleomorphic change (a decrease in hexagonal cell population) than enlargement of the mean cell size.Arc welding generates visible light and much ultraviolet (UV) light. Most of the UV radiation is absorbed by the superficial layers of the cornea. UV light of 265-275 nm wavelength causes photokeratitis (keratitis superficialis diffusa).' UV light of longer wavelength reaches the corneal endothelium, and that over 295 nm reaches the lens and can cause cataracts. However, there has been no report on the effect of UV radiation on the endothelial morphology of the human cornea.UV radiation is now thought to be one of the factors promoting cataracts with aging.23 The endothelium is also affected by aging.4-"' The mean cell size of the endothelium increases4'8 and hexagonal cells of the endothelium decrease with age.' "' The effects of aging on these changes are not clear. We assumed that UV radiation has an important role in the endothelial aging phenomenon and tried to study the relationship between the endothelial changes and UV radiation in welders, who are occupationally exposed to excess UV radiation.Corrcspondcncc to Dr 1. Karai. Subjects and methodsA total of 118 Japanese male welders of mean age 44-1 years (range 24-66 years) in-two welding factories in Osaka were studied. Their mean employment duration as welders was 17 4 years (range 1-39 years). The control group comprised 85 men of mean age of 41-8 years (range 20-73 years), employees in two chemical factories (42 males), and male outpatients (43 persons) without any corneal diseases from the Department of Ophthalmology in two hospitals. The welders were asked about their length of employment as a welder, the welding method used, and personal history of ocular disease. We excluded one welder with corneal disease, and one control who had once worked in welding. We also excluded persons with myopia gravis (over -6 dioptres) from this study.Specular microscopes (Keeler-Konan camera, and Bio-Optics) were used to photograph the central corneal endothelium of both eyes of the subjects three times after treatment with ophthalmic surface anaesthetic (0.4% oxybuprocaine hydrochloride).Kodak film (Tri-X pan, ASA 4...

The exposure-excretion relationship and possible health effects of exposure to methanol vapor were studied in 33 exposed workers during the second half of 2 working weeks. Urinary methanol concentrations were also determined in 91 nonexposed subjects. The geometric mean value for methanol in urine samples from the latter was less than 2 mg/l (95% upper limit of normal, less than 5 mg/l) when log-normal distribution was assumed. Among the exposed workers, the methanol level in urine samples collected prior to the work shift exceeded the 95% upper limit of normal. The time-weighted average intensity of exposure to methanol vapor was measured using personal sampling devices (in which water severed as an absorbent) in 48 cases of methanol exposure (i.e., 2 of the 33 exposed workers failed to provide urine samples, whereas 17 subjects were examined twice). Methanol concentrations in urine were determined in samples collected at the end of the shift from the 48 exposed cases as well as from 30 nonexposed controls. There was a significant correlation between the exposure to methanol vapor at concentrations of up to 5,500 ppm and the levels of methanol measured in the shift-end urine samples. The calculation indicated that a mean level of 42 mg methanol/l urine (95% confidence range, 26-60 mg/kg) was excreted in the shift-end urine sample following 8 h exposure to methanol at 200 ppm (the current occupational exposure limit). Dimmed vision and nasal irritation were among the most frequent symptoms complained during work. Three cases showing clinical signs of borderline significance were identified.

The validity of two new biological exposure markers of toluene in blood (TOL-B) and toluene in urine (TOL-U) was examined in comparison with that of the traditional marker of hippuric acid in urine (HA-U) in 294 male workers exposed to toluene in workroom air (TOL-A), mostly at low levels. The exposure was such that the geometric mean for toluene was 2.3 ppm with a maximum of 132 ppm; the workers were also exposed to other solvents such as hexane, ethyl acetate, styrene, and methanol, but at lower levels. The chance of cutaneous absorption was remote. Higher correlation with TOL-A and better sensitivity in separating the exposed workers from the nonexposed subjects were taken as selection criteria. When workers exposed to TOL-A at lower concentrations (< 50 ppm, < 10 ppm, < 2 ppm, etc.) were selected with correlation with TOL-A was examined, TOL-B showed the largest correlation coefficient which was significant even at TOL-A of < 1 ppm, whereas correlation of HA-U was no longer significant when TOL-A was < 10 ppm. TOL-U was between the two extremes. The sensitivities of TOL-B and TOL-U were comparable; HA-U showed the lowest sensitivity. Thus, it was concluded that TOL-B is the indicator of choice for detecting toluene exposure at low levels.