Document g4JOoyD4VbOz7yNqw0njMzXV

EVALUATION OF URINARY LEAD DETERMINATIONS* ' I. Th e Sig n if ic a n c e o s t h e Sp e c if ic Gr a v it y Le o Le v in e a n d Jo h n P. Fa h y Division of Occupational hygiene, Massachusetts Department of Labor and Industries, Boston, Massachusetts / "fTTHE determination of urinary lead is a tool 1 . of considerable importance in the control of industrial exposures to that metal. Much effort has been devoted to increasing the :liability of this tool through improvements in methods of sampling ,and( ^analysis,,and through l etter interpretation of analytical data. Relatively large scale field work such as occupies :iiis Division, especially -with respect to lead, the chief health hazard in Massachusetts industry, sets certain practical limitations to otherwise desirable methods of sampling. For example, we have found it impossible to arrange routinely for twenty-four l our samples, widely advocated as more reliable than instantaneous samples. Having accepted epot" samples, we at one time' hoped that it - ould be posable to note the time rate of secretion . > that results could be reported as mass per unit hie, as is also widely advocated. This, too, ; roved to be impractical. Hence for the past " ve years we have been taking spot samples and porting the results in terms of weight by volume ` ncentiation, i.e., milligrams per liter. Whatever defect may inhere in this method of apling and reporting, the findings have been `.rc-mely useful. Their general validity was abllshed by Elkins et al. (2) who showed high ^relation between urinary and atmospheric lings in 19 plants having lead exposures. Al- >ugh this work concluded that "measurement of 'Itr atmospheric or urinary lead will give a true >'ure of the lead hazard in the majority of in- trial processes," we have continued to make 'ii measurements, as a check, one on the other. rood correlation has continued, i --.riier Barnes investigated the "Possibilities ' Tirol of Lead Exposure by Examining Less 21 Hour Urine Samples" (3). His method study the variation in lead excretion by the Meek, Collins, and Harrold (4) made a similar investigation, analyzing 23 eight-hour samples which were composite parts of the nine 24-hour samples. The standard deviation of the spot . sample findings from the 24-hour finding was 0.026 mg. per liter. It was concluded that the percentage error for spot samples was large for low lead samples (normal range) but satisfactory on urines indicating absorption. Notwithstanding the general validity and estab lished utility of urinary lead determinations, In cluding spot samples, they have been subjected to criticism for which remedies have been and should be sought; This criticism is based on the fact that urinary lead concentrations are not as constant as blood lead levels over short time intervals in cases of constant exposure or mild or advanced plumbism. According to Kehoe, et al. (5), reporting on results of lead ingestion by human adults, . .lead concentration in the blood of the. subjects was subject to insignificant variation during the course of a day, and to only minor variations from day to day, while the urinary lead concentration, as well as output per unit of time, varied widely in accordance with the volume of water excreted by the kidneys.!' Thus stated, the criticism applies not only to spot samples, but also to 24-hour samples, and to findings based on the time rate of lead excretion. If this is tiie case, the evaluation of spot samples on the basis of how they compare with 24-hour samples, carried out by Barnes, Meek, and others, afford only limited comfort. Indeed, Barnes (3) warns that "it must be borne in mind that 24or 48-hour specimens from the same person show quite a wide range of lead concentration and of lead excreted per day..." Webster (6) believes that "the total 24-hour urinary lead output has more significance than any other measure of lead made on fractional-day exposed subject from whom several spot t had been obtained in two days. The -on was sufficiently low to satisfy him that mples were adequate for the purpose. samples," and finds against spot samples whether reported* as lead concentration or time rate of excretion. Attempts have been made to correlate urinary wed for publication May 18, 1945. lead findings with other properties of the urine 217 N36924 218 JOURNAL OF INDUSTRIAL HYGIENE AND TOXICOLOGY [vol. 27, no. f in the hope that this would provide a basis for solids. This would be a very simple matter to do, reporting or correcting findings, resulting in since total urinary solids is directly proportional diminished variation. Webster found considerable roughly, to the specific gravity minus unity (7). diurnal variations, unrelated to specific gravity, This relation is discussed in greater detail below. pH, time, volume, and blood-lead content. He found some correlation with total phosphate in the n urine. It is the purpose of this paper to present data Barnes (3) looked for, but failed tofind, a correla winch constitute evidence of the following: tion between lead concentration on the one" band, 1. That the specific gravity is an important and specific gravity, acidity, nitrogen, and total property of a urine specimen from the viewpoint volume, on the other. He did find a decrement in of evaluating urinary lead excretion and that its variability when results were calculated in y per measurement can and should be integrated with hour. Pinto and others (1), using similar data, the analytical findings. studied the variability of findings from individuals 2. That when the specific gravity is properly so recovering from plumbism, calculating the results integrated, the instantaneous, or spot sample is at in three different ways for comparison, as follows: least as valid as the 24 hour sample calculated (1) as r/100 ml., (2) as y per 100 ml. times the conventionally. square root of the volume excreted per minute, and 3. That when an individual with an abnormally (3) as y per hour. They found that the second high, fairly constant blood lead level (a condition i\ method gave the lowest coefficients of variability. which obtains for long periods during recovery Urinary findings were sufficiently constant for from plumbism) is subjected to a series of lead i i them to deduce an empirical formula giving the urinalyses, greater constancy of excretion is found approximate blood-lead level in terms of the when the results are adjusted for the specific urinary finding. The standard deviation of the gravity of the samples, than when calculated either ratios urinary lead to blood lead ryas 36 per cent as weight per unit volume, or as weight per unit J of the mean. time. The constancy is approximately equal to j \ Unfortunately, however, it was not practical to that found when calculating by the volume-time integrate the results of this work into our routine 1 studies because it required notation of the time s '} interval during which each sample was secreted. 4. That the mass of lead excreted per unit mass Thiswould negate the advantages of spot sampling. of total urinary solids is more constant, under . Kchoe and co-workers (5), in their studies of conditions where the blood lead level is fairly lead ingestion cited above, showed a high negative constant, than the mass of lead excreted per unit t .1! correlation between the volume of urine excreted of volume of urine, or per unit of time. atid the lead concentration:--They stated, "The urinary lead concentration.. .varied over a wide , HI !t range during each 24diour period, and the plotted This laboratory has appreciated the importance curves of these concentrations are almost exactly of reporting the specific gravity concomitantly with 1 opposite, throughout their course, to the corres results of urinalysis. The specific gravity reading 1 ponding curves of urinary volume. It is apparent has therefore always been taken and noted in our .. .that no factor other than the volume of available reports to medical and industrial establishments. j. water exerted any important role in the induction of This practice is, in fact, quite general. But it may variations in urinary lead concentration during well be asked what use is made of this information. these 24-hour periods." (Our emphasis.) It would seem that a certain allowance is made If the volume of available water is responsible mentally for the specific gravity in evaluating the 4 for the variations in lead concentration, it is also lead finding. For example, a distinction is made responsible, In whole or part, for variations in the mentally between a finding 0.2 mg. per liter on a concentration of other normal constituents, of the sample of gravity 1.010 and the same finding on urine, for example, total dissolved solids. This one of 1.030. It is recognized that the former { raises the interesting question whether greater may well be more serious. But however valuable constancy could be expected if the results were this supplementary information may be, it must calculated as mass of lead per unit mass of total be admitted that its use constitutes reliance on * i DUP050312872 MS] URINARY LEAD DETERMINATIONS. I 219 , ^;-hly subjective kind of mental reservation or iliowance. An interesting illustration of the kind of sub: ;tive "correction" that is relied on in evaluating anJings is contained in the remark of Meek and , workers (4), viz.: "Previously Barnes (3) has ported good results using instantaneous samples, :.-,r:icularly when the specific gravity of the sample _ j s noted." (Our emphasis.) It does not seem ;jus that the results themselves were improved .imply by noting the specific gravities, since no other use was made of them. Now, if it be conceded that the specific gravity cues have some bearing on the results, and that attain modifications, mental or otherwise, are in subtracting 0.0001 for each degree F. above or. below 60 degrees. However, we did not correct our readings for temperature. The majority of" our readings are probably correct to within 0.002. The samples were analyzed by the dithlzone titrimetric extraction of the oxidized calcium' oxalate eoprecipitate (9). Our object was to determine whether or not .a. correlation existed between the specific gravity and' the lead content A table was constructed for each year, noting the lead content of each sample under its specific gravity. As would be anticipated, the number o samples falling in the various specific gravity 1 No. of Samples 167 147 161 145 266 137 134 TABLE 1 Sp e c if ic Gs a v it v a n d Le a d Co n t en t , 1157 Sp o t Sa mpl e s 2 345 67 8 ?G Specific Grav ity Class G (Average Gravity) U (Ave. *b) 7/100 ml. 'V G va (Ave. Pb corrected for gravity) (SamVeousing Eq. 2) . v* (Ave. Fb, In y/gram solids). 3-16 17-20 21-23 24-25 26-28 29-30 31-40 12.5 19.0 22.0' 24.5 26.9 29.7 32.7 9.4 12.9 14.0 17.4 18.6 18.719.9 0.75 0.68 0.64 0.71 0.69 0.63 0.61 17.S 16.2 15.1 16.9 16.5 15.0 14.5 17.0 16.1 15.2 16.9 16.5 14.8 14.6 2.89 2.60 2.44 2.73 2.65 2.42 2.34 Standard Deviation .... Coefficient of Vari- ability,.. 23.8 6.52 27.4 15.9 3.55 22.4 0.67 0.058 6.8 16.0 1.03 6.4 15.9 0.93 5.8 2.58 0.175 ler contingent on their notation, then it is worth while to try to ascertain as precisely as may be just hat that bearing is and how it can be integrated `.o the result itself. in an effort to do this, we undertook an analysis the accumulated data for the past three years " h-ad urinalyses of spot samples. A total of :[57 samples were run in this period. In every - e there was at least some basis for suspicion of i exposure, and in the overwhelming majority ; cases there was, in fact, greater or less exposure, 'e data for each sample consist of its specific - and its lead content in micrograms per 51 ad- The specific gravity was determined with 1 ' ! ating spindle urinometer graduated for use F. The correction involves adding or columns tended to form the curve of normal distribution, with the mode the same for each year, namely at G = 26 (i.e., specific gravity = 1.026. This notation will be used henceforth). In the vicinity of the extremes, G = 3 and G = 40, there were naturally too few samples for a valid mean to be taken, and classes of gravity were established so that about 50 samples would fall into each class per year. For the three years, each gravity class would have roughly 150 samples, a number that would provide a satisfactory average lead content figure. The average lead content of the samples in each gravity class is shown'in Tabic 1, Column 4. It is seen that for each class the average lead concentration increases as the average specific gravity increases. This is shown graphi- DUP050312873 220 JOURNAL OF INDUSTRIAL HYGIENE AND TOXICOLOGY 27, i cally in Figure 1. The correlation between the two averages was calculated by Pearson's formula for the correlation coefficient, r: 0 0(G - CTj) - Gi)+(U- Ut) - G2) . . ~t~ ~KU -- G,,) (G -- G,,) N r -- ----------------------------------- :---------------------------<rtj<ro . r = -j-0.94 Perfect correlation is r -- +1. The high correlation of + 0.94 between the aver age specific gravities and lead contents of seven groups of spot samples classified by specific gravity, embracing an aggregate of 1157 samples largely obtained and those under which in all probabilitv the clinical or hospital data were compii;-,: Practically all our samples were taken trom workc; who were on the job during the secretion of th. sample. The increased metabolism of persons a: work would be expected to result in increased rcr,:,! excretion of solids and in higher urinary conccntra tions of solids. Mean urinary lead concentrations in y/100 nil. (G<?) corrected for gravity by Equation (1) \vcre calculated for the seven classes of samples and appear in Column 6, Table 1. It is seen that the coefficient of variability of the corrected value- is 6.4 per cent, that for the uncorrected is 22.4 per cent. H r-1 rl r-5 . SPECIFIC ORAVITX Pto. 1. Specific gravity and mean lead content, 1157 samples. from exposed workers, is evidence that the specific gravity influences urinary lead concentrations and should be integrated with the findings. We have considered three ways in which the specific gravity may be applied in calculating results. The first and simplest is to calculate all results to the mean gravity of all samples, G = 24. That is, The ratios of the mean lead concentrations to t' mean gravities fg y of the several classes y shown in Column 5, Table 1. This ratio isfairly ci ' stant, the coefficient of variability being only 6.5 r cent. Application of the mean of tins ratio, 0.* ` in the following equation is a second method integrating the specific gravity with the urinal> -; 24 (1) Go = G -- (2) If G < 24, Go = G +0.67.(24--G) If G > 24,Go = G --0.67 (G--24) Perhaps a few words should be said about the fact that the mean gravity of our spot samples is considerably higher than mean normal urinary gravities cited in the medical literature. We believe this is adequately explained by the con trasting conditions under which our samples were Column 7 gives the corrected results cala:!'.' in this manner. The third method of making the correction - calculate the result as mass of lead per unit " of total solids. Total urinary' solids may be roughly esti--` i ,wr.?rfr.;-Tr DUP050312874 Oil. 1945\ URINARY LEAD DETERMINATIONS. I 221 from the volume and the specific gravity by appli cation of Lang's formula, 5 = 2.6 G (7) where S is total solids in grams per liter. Then ,, 10 U 3.S U Vs = 2.6 G " G ' v.-hcre Us is micrograms of lead per gram of solids (or parts per million), and U is the urinary lead concentration in micrograms per 100 ml. It is evident that the value Us is directly proportional to Vo of Equation (1), for Us 00 (/ ~G~ The values for Us for the averages of the seven groups of samples appear in Column 8, Table 1. The reduced variability coefficient compared with tJ is evidence that lead excretion calculated on the basis of total solids is more valid than when calculated on the baas of urine volume. There is every indication that the improved validity which many investigators have obtained with 24-hour samples as compared with spot samples is due pri marily to the fact that the 24-hour sample tends, of course, to approach the mean concentration of urinary solids, or what is substantially the same thing, tends to approach the mean specific gravity. But if the spot sample is adjusted to mean specific gravity, or calculated on the basis of its solids, then it should have a validity equal to or greater than the conventionally calculated 24-hour sample. Data specifically confirming this are not'available, but it appears to us to follow as a corollary of the demonstrated correlation between specific gravity and lead concentration. The high negative cor relation found by Kehoe and co-workers (5), referred to above, between the volume of urine excreted and the -lead concentration, is a further indication that "the .volume of available water" can induce lead variations in 24-hour samples that will be compensated for by adjusted spot samples. The foregoing three methods of correcting for specific gravity are equivalent, as may be noted uom the tabular values and from the nearly equal coefficients of variability. Probably the first method (Equation 1) is to preferred because of its simplicity and because it involves the least radical departure from con ventional methods of reporting results. A word should be said about samples of -excep tionally low specific gravity. . The question will arise whether it is valid to triple, or more than triple the lead concentrations found in samples of G < 8. Only 20 samples out of 1157 had specific gravities of 1.008 or less, and their mean lead concentration was 0.058 mg. per liter. Li our opinion, the only consideration that militates against making the adjustment to mean gravity is the fact that the analytical precision drops sharply as the lead concentration of a sample decreases. The sensitivity of any method is a fixed quantity, say 1 microgram. For a 50 ml. sample having 0.2 mg. lead per liter the probable error is 10 per cent. For the same size sample containing 0.05 mg. per liter, the error is 40 per cent. The adjustment would obviously not be justified in this case. However, the precision may be increased by taking three or four times the usual quantity of sample for analysis. The method of adjusting the concentration to mean specific gravity (Equation 1) was used in the calculations involving the data presented below as additional evidence of the validity of this adjust ment. IV It has been assumed that the best test as to the reliability of the basis for calculating and reporting results of urinary lead determinations is the degree of constancy that can be shown under conditions where constancy is to be expected, namely, where the exposure is constant, or where the blood lead concentration is constant, or where recovery from plumbism is occurring by the gradual excretion of lead stored in the tissues. The appeal to this criterion is general throughout the literature on this subject. Under such conditions of expected constancy, there is no apparent reason for wide variations in renal lead excretion within relatively short time intervals, except where there is impair ment of renal function. The evidence is that such reported variations are due to fallacies inherent in the determination which can be eliminated by fur ther study. It is true that other reasons for variations have been suggested. Pinto and' co-workers (1), seeking a relation between urinary and blood lead, have said that "since only 10 per cent of the total lead excreted by the body is eliminated through the kidneys, the question can very well be raised as to V:?'V l s ! i I .5 222 JOURNAL OX-' INDUSTRIAL HYGAViE AND TOXICOLOGY [vol. 27, r.., how significant urinary lead determinations may be." However, Kehoe and co-workers (8) in their later report on lead ingestion experiments state ",.. that lead evacuated from the alimentary tract is largely lead which has been swallowed, and that the true alijnen tary excretion of lead is small. A large proportion of the lead actually absorbed into the body is excreted by way of the kidney. Certainly 'the renal lead excretion is quantitatively as great as the alimentary, and probably it is greater, On this point, clearly, the generally accepted idea of the means by which lead is excreted will have to be revised." This is the highest tribute yet paid to the significance of urinary lead excretion. The work just cited also considerably weakens the theory that dietary factors may induce wide variations in renal lead excretion. Summarizing the effects of induced dietary changes the authors express "grave doubt" that the normal, spontane ous elimination of the toxic agent can be "materi ally accelerated by artificial means." And they conclude that their tests show that .. within the limits of lead absorption dealt with..., the lead metabolism is exceedingly stable,--so stable, in fact, that it is not importantly influenced by gross changes in the add-base equilibrium or by the cal cium and phosphorus metabolism of the body." This test of constancy was applied to the data for ten series of samples taken from ten subjects, three of whom (data of Bames (3)) were spray painters with a presumably constant lead exposure of long duration, and seven of whom (our data) were recovering from lead poisoning. The scries for each individual consisted of six or seven sam ples (in one case, five) collected over a' period of about 12 to 48 hours. The lead excretion was calculated by two methods, simply as concentra tion in y per 100 ml., and as y per 100 ml. corrected - 24 to mean specific gravity by the factor It Cr should be recalled that the latter method is the equivalent of reporting the result on the basis of mass of lead per unit mass of total urinary solids. The variation within each series for each method was evaluated as the coefficient of variability. This term is simply the standard deviation of each series expressed as a per cent of the mean and is the mathematical measure of relative variation, or dis persion about the mean. Tabic 2 shows the co efficients of variability for the two methods. It is seen that by integrating the specific gravity into tl-.i i-evjlts, the variability decreased in eight serf - '-.creased in two. There was a mean decree ir. variability of 13 per cent. ijirees calculated the results of his three sen (A, Ji, and C, Table 2) as micrograms per hour. Hr. tv et al. (1) calculated our1 data both as micr, per hour and- in terms of volume per uid: TABLE 2 Co2.3-rrciF.x-TS o f Va r ia b il it y o f 10 Se r ie s Ur in ar y Le ad De t er min at io n s o; j c k ;k c t NO. OF 3A2CTI.ES ANALYSIS RETORTED AS y/lQO NX. TINAD JUSTED FOR SPECIFIC GRAVITY ANALYSIS RETORTED AS 7-/J00 1TL., ADJUSTED FOE SP. GS. BY FORMULA PER CENT d ec eeas v IN VARU. BtLIT'i A 6 3619 20.8 43.6 B 6 31.0 43.1 -39.0 C 7 28.3 24.3 14.1 M S 37.0 35.3 4.6 N 7 19.6 14.3 27.(1 O 7 28.9 21.6 25.2 V 7 47.0 39.0 17.0 Q 7 27.2 23.6 13.2 R 6 29.8 30.1 -1.0 S 6 35.8 27.0 24.6 Mean... 32.2 27.9 13.0 TABLE 3 Me a n Co e f f ic ie n t s o f Va r ia b il it y o f For: ME-riions o f Ca l c u l at in g .Ur in ar y Lead De t e r min a t io n s . V, t /h r . (10 SUBJECTS) (11 s u b j e c t s ) 32.2 34.5 ir-vA Vt (8 SUBJECTS) 27.1 vg <10 SUBJECTS' 27.9 time by the formula U Thus we have a bi ': for comparison of the variability of four methods < calculating the results. Table 3 shows the mecoefficients of variability of the four methods. will be noted that the method of adjusting rew-' to mean gravity (Vg) and the method based on t" formula show approximately cquival . " variability, and effect a decrease of about 13 1 ' cent compared with the result as a simple com'' li.e., data of this Division, under whose auspice? ' cited work was carried out. ygay-.... . DUP050312876 * <b M^ibsC Orl. 1S4S[ URINARY LEAD DETERMINATIONS. I 223 tration. The result expressed as micrograms per hour shows no decrease in variability, confirming the finding of Webster (6). Su mma r y 1. A high correlation was found between specific gravity and lead concentration for 1157 spot urine simples obtained largely from workers with more or less exposure to lead. 2. A decrease in variation of urinary lead excre tion of series of spot samples taken from subjects recovering from lead poisoning was found when the results were adjusted to mean specific gravity, as compared with the unadjusted concentration, or with results expressed as time rate of lead excretion. 3. These findings constitute evidence that renal lead excretion calculated as mass of lead per unit mass of total urinary solids is more representative of true exposure than when calculated as mams per unit volume or per unit time. 4. Calculated in this manner, the evidence is that spot samples are not less representative than 24-hour samples calculated in terms of weight per unit volume. 5. It is recommended, therefore, that urinalyses for lead be reported as "milligrams per liter ad justed to mean specific gravity", a figure obtained by multiplying the analytical finding in mg. per 24 liter by the factor --, where G represents the sig- (j riificant figures of the specific gravity, i.e. G - 1000 (sp. gr. -- 1). Grateful appreciation for reviewing this paper is expressed to Dr. Joseph C. Aub, Massachusetts General Hospital, Captain Hervey B. Elkins, Chemical Warfare Service, Army of the United States, and Dr. Robert A. Kckoe, Director, Ket tering Laboratory of Applied Physiology. BIBLIOGRAPHY (1) Pin t o , S. S., El k in s , H. B. a n d Eg e, J. F., Jr .: Renal excretion of industrial chemicals. I. Urinary-lead concentration and blood-lead clear., ance. Tins J., 23: 7, pp. 313-321, Sept., 1941. (!.' El k in s , H. B., Eg e, J. F., Jr . a n d Ru o t o l o , B. P.: Evaluation of the lead hazard. Urinary vs. atmospheric lead. Th is J., 23: 6, pp. 256-258, June, 1941. !->) Bar n es , E. C.: Possibilities of control of lead ex posure by examining less than 24 hour urine samples. Th is J., 21: 9, pp. 464-468, Nov., 1939. ' i; Me e k , S. F., Co l l in s , G. R. a n d IIa r r o l d , G. C.: Correlation coefficient between basophilic aggre gation test and lead in urine. Tars J., 22: 9, pp. 401-407, Nov., 1940. u Ke iio e, R. A., Ch o l a x , J., Hu b ba r d , D. M., Ba mb ac h , K., Mc Na r y , R. R. a n d St o r y , R. V.: Experimental studies on the ingestion of lead compounds. Th is J., 22: 9, pp. 381-400, Nov., 1940. (6) Webs t er , S. H.: Diurnal variation of urinary lead excretion. Public Health Reports, 56: pp. 1834-1848, Sept. 12,1941. (7) Bo d an s k y , M.: Introduction to physiological chemistry, p. 381, John Wiley & Sons, Inc., 1930. (8) Keh o e, R. A., Ch o l ak , -J., Hu b ba r d , D. M., Ba mb a c h , K. a n d Mc Na r y , R. R.: Experimen tal studies on lead absorption and excretion and their relation to the diagnosis and treatment of lead poisoning. Th is J., 25: 2, pp. 71-79, Feb., 1943. (9) Ross, J. R. a n d Lu c as , C. C.: A new method for the determination of minute amounts of lead in urine. Jour. Biol. Chem., Ill: 2, pp. 285-297, Oct., 1935. i J' l b It. ; *fi, ,i r' v; f DU P050312877