DiscussionThe present study makes it clear that an HPLC methodcan be applied for determination of toluene in urine. Themethod can analyze about 40 samples (excluding internalquality-control samples) daily when the analytical instrumentis allowed to run 24 h/day. It should be stressedthat the overnight operation was permitted for the HPLCsystem because there was no use of any explosive gassuch as hydrogen in the system, conditions which areconvenient from viewpoints of work performance andwork safety, especially when mass analysis is desired.Other advantages of the present HPLC determinationover the HS-GC method (Kawai et al. 1996) include theneed for fewer manual procedures for sample pretreatment,i.e., just the addition of a 1-ml urine sample to anHPLC bottle containing 1 ml acetonitrile and 100 llmethanol, followed by sealing and gentle mixing. Duecautions should be exercised, however, to prevent possibleloss of toluene from urine or contamination ofurine samples with toluene in ambient air during theprocedures from the time of urine collection to that oftransfer of an aliquot to the bottle (Kawai et al. 1996).Such simple procedures will contribute substantially tothe removal of analytical errors.The co-presence of lipophilic acetonitrile in theHPLC bottle enables the storage of urine samples formuch longer periods without urine spoilage as comparedwith HS-GC samples. Experience (Fig. 2) shows that theHPLC samples can be stored for more than 1 week whenkept at 4 °C versus a shorter period in the case of HSGCsamples (Kawai et al. 1996); a longer storage life is asubstantial advantage when the method is applied to thepractice of occupational health service.No explanation is currently available as to why toluenerecovery was reduced with urine samples fromuntreated diabetics (Table 1) and why methanol removessuch suppressive effects. The choice of methanolas the agent was made just by chance and not by theoreticaldeduction. Nevertheless, one should be awarethat diabetics without apparent subjective symptomsTable 2 Toluene in urine versus toluene in breathing-zone air; the correlation with or without correction for urine densityaCase correction for Unitb Regression parametersa b r PA. With data as observedWith a combination of 13 exposed and 10 non exposed workers:None (i.e., as observed) lg./l 4.34 0.73 0.97 <0.01Creatinine lg./g creatinine 2.55 0.50 0.94 <0.01Specific gravity (1.016) lg./l 3.82 0.41 0.85 <0.01With 13 exposed workers only:None (i.e., as observed) lg./l 20.90 0.56 0.91 <0.01Creatinine lg./g creatinine 12.31 0.40 0.84 <0.01Specific gravity (1.016) lg./l 18.42 0.26 0.56 <0.01B. With data after double-logarithmic conversionsWith a combination of 13 exposed and 10 non exposed workers:None (i.e., as observed) lg./l 0.26 0.63 1.00 <0.01Creatinine lg./g creatinine 0.23 0.75 0.99 <0.01Specific gravity (1.016) lg./l 0.22 0.72 0.99 <0.01With 13 exposed workers only:None (i.e., as observed) lg./l 0.66 0.62 0.95 <0.01Creatinine lg./g creatinine 0.35 0.69 0.88 <0.01Specific gravity (1.016) lg./l 0.14 0.61 0.81 <0.01a Values in the table are parameters of the calculated regression lineY a + bX, where X is the TWA toluene concentration (in ppm)determined in breathing zone air and Y is the toluene concentrationdetermined in end-of-shift urine samples (units are shown in thetable). The case of the 13 exposed workers and the 10 non exposedworkers in combination and that of the 13 exposed workers onlyare shown separatelyb Units for urinary toluene concentrationca, b, r and P are the intercept on the vertical axis, the slope, thecorrelation coefficient, and its statistical significance, respectivelyd Regression analysis was conducted using log10 X and log10 Y, inwhich 0 ppm toluene in air and 0 lg. toluene/l urine were replacedwith 0.5 ppm and 1 lg./l, respectively, both being one-half the detectionlimit306may well be left untreated, as was the case for subjects Cand D. The confounding effects of an insidious diseasesuch as diabetes mellitus on biological exposure monitoringby urinalysis well deserve further study, especiallyon the mechanism of the interference in urinalysis.Although the number of toluene-exposed workersexamined in the present study was limited, a preliminarystatistical analysis was conducted to assess the lowesttoluene concentration at which the exposed group couldbe statistically separated from the non exposed subjects(or the LSC-1 value; Kawai et al. 1992). Statistically, theLSC-1 value can be expressed as the toluene exposureconcentration (thus, in ppm) at which the lower 95%limit, e.g., of urinary toluene concentration, is equal tothe upper 95% limit of the group mean of, e.g., urinarytoluene concentration, for those without exposure (i.e.,the upper 95% limit of the background level; Kawai et al.1992).When the proposed procedure was applied to thepresent results, it was found that the upper 95% limit ofthe group mean at zero ppm toluene concentration (i.e.,without toluene exposure) was equal to the lower 95%limit after toluene exposure at 10±15 ppm (Fig. 4A).This value of 10±15 ppm as the LSC-1 is roughly equal tothe value obtained when urine samples from more than100 workers were analyzed by the HS-GC method(Kawai et al. 1996), although the former value should betaken as preliminary due to the limited number of casesanalyzed. A similar analysis of the results presented byMonster et al. (1993) on toluene in urine samples collectedat 16 h after the termination of toluene exposuregave an LSC-1 value of 20±25 ppm. Hence, it would beprudent to conclude that urinalysis for toluene is probablysensitive enough to allow statistical separation ofthe workers exposed to toluene at the occupationalexposure limit of 50 ppm (American Conference ofGovernmental Hygienists 1996; Deutsche Forschungsgemeinschaft1996; Japan Society for OccupationalHealth 1996) from the non-exposed subjects.In practice, the time of urine sampling is an importantfactor in evaluation of the analysis result. Whereasthe end of the afternoon shift was tentatively selected asthe time of sampling in this preliminary study, byanalogy to the sampling time for hippuric acid (Ikeda1996), it may be necessary to identify the best time ofurine sampling for toluene to obtain the most representativemeasures, taking the kinetics of urinary tolueneexcretion into consideration.The account of toluene to be excreted into urine astoluene itself is of theoretical interest. This was calculatedwith three assumptions: that the respiratory volume underthe working conditions was 10 l/min, that the rate forthe absorption of toluene in the lungs was 50%, and thatthe urine volume was 1 ml/min (Yasugi et al. 1994). Theexposure to toluene at 50 ppm (187.5 mg/m3) will inducetoluene absorption in the lungs amounting to 937.5 lg/min (187.5 mg/m3 ´ 10)2 m3/min ´ 50/100). Given theregression parameters from Table 2, i.e., the interceptand the slope are 4.34 lg./l and 0.73 lg. l)1 ppm)1, theexposure to toluene at 50 ppm will result in the urinaryexcretion of toluene at 40.84 lg./l [4.34 + (0.73 ´ 50)],or 0.0408 lg./min. Thus, the excretion/absorption ratio is
0.041/937.5 0.00004, or 0.004%.
Similar calculations taking regression lines of
Y
mg/l 270:7
mg/l
31:1
mg lÿ1 ppmÿ1
X
ppm
for hippuric acid and
Y
lg/l 362:1
lg/l
11:1
lg lÿ1 ppmÿ1
X
ppm
การแปล กรุณารอสักครู่..
