Objectives: This study was initiated to make comparative evaluation of five proposed urinary markers of occupational exposure to toluene, i.e., benzyl alcohol, benzylmercapturic acid, omicron-cresol, hippuric acid and un-metabolized toluene.
Methods: In practice, six plants in Japan were surveyed, and 122 Japanese workers (mostly printers; all men) together with 12 occupationally nonexposed control subjects (to be called controls; all men) agreed to participate in the study. Surveys were conducted in the second half of working weeks. Time-weighted average exposure (about 8 h) to toluene and other solvents were monitored by diffusive sampling. End-of-shift urine samples were collected and analyzed for the five markers by the methods previously described; simultaneous determination of omicron-cresol was possible by the method originally developed for benzyl alcohol analysis.
Results: The toluene concentration in the six plants was such that the grand geometric mean (GM) for the 122 cases was 10.4 ppm with the maximum of 121 ppm. Other solvents coexposed included ethyl acetate (26 ppm as GM), methyl ethyl ketone (26 ppm), butyl acetate (1 ppm) and xylenes (1 ppm). By simple regression analysis, hippuric acid correlated most closely with toluene in air (r = 0.85 for non-corrected observed values) followed by un-metabolized toluene (r = 0.83) and o-cresol (r = 0.81). In a plant where toluene in air was low (i.e., 2 ppm as GM), however, un-metabolized toluene and benzylmercapturic acid in urine showed better correlation with air-borne toluene (r = 0.79 and 0.61, respectively) than hippuric acid (r = 0.12) or o-cresol (r = 0.17). Benzyl alcohol tended to increase only when toluene exposure was intense. Correction for creatinine concentration or specific gravity of urine did not improve the correlation in any case. Multiple regression analysis showed that solvents other than toluene did not affect the levels of omicron-cresol, hippuric acid or un-metabolized toluene. Levels of benzylmercapturic acid and un-metabolized toluene were below the limits of detection [limit of detections (LODs); 0.2 and 2 microg/l, respectively] in the urine from the control subjects.
Conclusions: In over-all evaluation, hippuric acid, followed by un-metabolized toluene and omicron-cresol, is the marker of choice for occupational toluene exposure. When toluene exposure level is low (e.g., 2 ppm), un-metabolized toluene and benzylmercapturic acid in urine may be better indicators. Detection of un-metabolized toluene or benzylmercapturic acid in urine at the levels in excess of the LODs may be taken as a positive evidence of toluene exposure, because their levels in urine from the controls are below the LODs. The value of benzyl alcohol as an exposure marker should be limited.