Absence of toxic effects in silver reclamation workers

toxiceffectsinsilverrecla mation workers. Scand J Work Environ Health 1989;15:210-221. Recentreports haveallegedthat silver presentsa toxic hazard to exposed workers. To define the potential risks of long-term exposure to silver better, a cross-sectional investigation was conducted of 27 Caucasian males occupationally exposed to primarilyinsolublesilvercompounds and 27 matched referents. Physicalexamination and electronmicro scopy of skin biopsies revealed no cases of generalized argyria. Measurements of facial discoloration judged from color photographs by panels of laymen and physicians, showed no significant difference betweenthe two groups. Although 29 % of the silverworkers and none of the referents exhibited ocular silverdeposition, optometric and contrast sensitivitytest results revealedno significant deficits in visual performance. The kidneyand respiratory findings wereessentiallynormal in both populations. Despite the increasedpresenceof silverin the blood, feces, and hair of the recoveryworkers versusthe referents, there was no evidencethat chronic silver exposure adverselyaffected the health of these employees.

From an occupational health standpoint, interest in the dermatologic effects of silver (argyria) has lessened considerably during the past few decades as changes in manufacturing processes and improved industrial hygiene practices have significantly reduced workplace exposures . Toxicologic evidence has generally supported the clinical observation that chronic exposure to metallic silver and its soluble salts (and pre sumably insoluble compounds) does not appear to represent a serious health risk to man. However, health data from clinical studies in the United States (1)(2)(3), and case histories reported in the European literature during the past 15 years (4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15) have suggested the need to reevaluate the human toxicologic effects of silver for such target organs as the skin, eyes, lungs, and kidneys. In addition, it is important to define more precisely the potential risks of long-term exposure to insoluble forms of silver, which have been incompletely studied and for which there is no workplace standard in the United States (US). This paper describes air sampling results and clinical and epidemiologic findings for a group of silver reclamation employees and their matched referents at a large photographic manufacturing facility . Specific study objectives included investigations of (i) argyria (localized and generalized), (ii) argyrosis (ocular sil- (17). In 1971 the US Occupational Safety and Health Administration (OSHA) established a standard of 10 ug Ag/m' for both silver metal and its soluble compounds (18). Finland and the United Kingdom have promulgated standards of 10 ug Ag/m' and 100 Ilg Ag/m ", respectively, for soluble and insoluble forms of silver (19).
The estimates of the potential annual average for personal exposure rates (8-h basis) shown in table I were determined from an analysis of more than 100 area and personal (breathing zone) samples collected since the mid-1950s. Samples of airborne dust were collected on mixed cellulose ester filters and analyzed by either atomic absorption spectrophotometry or inductively coupled plasma atomic emission spectrophotometry. Silver exposures in this environment are highly variable due to the extreme values (thousands of micrograms per cubic meter) which may occur for limited periods during cleaning and maintenance operations and during nonroutine events such as equipment malfunctions. In the determination of annual occupational exposure rates, the average of full-shift sampling data for an occupation was increased on the basis of an estimate of the frequency of irregular, short-term, highexposure tasks. Historically, the number and severity of such exposures have varied with the introduction of new procedures, often compounded by equipment problems. Because of individual work practices, including the use of respiratory and facial protection, and differences in occupational assignments, actual personal exposures may have differed from the values in table I, which should be considered as approximations of chronic silver expo sure.

Subjects
Thirty-one men with five or more years of experience (by 1979) in those areas with the highest potential for exposure were eligible to participate in the study. Included were burner operators, maintenance workers, mechanics, and smeltery and centrifuge operators; four employees were supervisors. Because of the difficulty in diagnosing argyria, three nonwhites were excluded from the analysis, although they were given standard clinical examinations, including laboratory studies. In addition, one worker chose not to participate. The final exposed group, therefore, included 27 Caucasian men.
An equal number of employees , matched with the exposed subjects by age (5-year category), sex (male), and race (Caucasian), were randomly sampled from one manufacturing (N = I I) and two office (N = 16) departments in which silver was not used. None of the referents had an occupational history of silver exposure. Since participation in the study was voluntary, provision was made for the random selection of alternates.
Using the same criteria, a physician and an occupational health nurse shared responsibility for examining the study subjects during February-April 1979. Optometric examinations were conducted in January 1981. The clinicians were blinded as to the employee's occupational status (exposed or reference subject).

Health questionnaire
The subjects completed a brief, self-administered medical history form which requested information about smoking habits, alcohol consumption, medication usage, and respiratory symptoms and illnesses.

Physical examination
In addition to height, weight, and blood pressure measurements, the clinicians examined each employee's face, neck, nasal and oral mucosa, and gums for signs of argyria.

Facial photograph study
The objective of the facial photograph study was to determine whether the silver workers had more facial discoloration than their matched referents and, if so, whether it was judged to be cosmetically objectionable. Panels of laymen and physicians evaluated color photographs of the faces of each matched pair of study subjects for the presence of skin discoloration. The portrait judged to have the most pigmentation was rated according to the degree of cosmetic objectionability. (A description of the study methods is included in the appendix.)

Electron microscopy
The subjects included five randomly selected silver workers with an average of more than 29 years of recovery experience and two referents. Skin biopsies from the buttock region were assessed for the presence of silver by microscopic X-ray diffraction analysis .
(The laboratory techniques are discussed in the appendix.)

Ophthalmologic studies
Twenty-four silver workers and 22 referents were evaluated in 1981 by an experienced optometrist. Each was given a standard optometric evaluation, including visual acuity (uncorrected and corrected), Schiotz tonometry, keratometric reading, and an ophthalmoscopic examination of the fundi. The data were analyzed separately for the right and left eyes, and in the case of visual acuity also for binocular viewing. In addition, conjunctival pigmentation was graded by direct visualization, while the cornea and lens were examined by slit-lamp biomicroscopy. As has already been indicated, all the studies were conducted without knowledge of the employee's occupation.

Contrast sensitivity experiment
Contrast sensitivity is a measure of the eye-brain capacity to detect luminance contrast at various spatial frequencies. It is determined by changing the contrast (light-dark difference) of a spatial sine wave target around a threshold level at a selected frequency until an image is perceived P 0J0 of the time . Performed in conjunction with standard visual acuity testing, this technique offers potential value in diagnosing ocular impairment in a wide range of clinical situations (20)(21)(22).
During 1981, an experiment was conducted to test the hypothesis that the deposition of metallic silver in the cornea might act to scatter incoming light in the ocular media and thereby cause a retinal image of reduced contrast, particularly under veiled glare conditions (23,24). Specifically, if silver deposition were responsible for a significant degree of retinal contrast loss, the recovery workers would be expected to exhibit lower contrast sensitivities than the referents. Furthermore, it was hypothesized that these values would be differentially reduced under conditions of glare illumination.
Of the 54 individuals originally studied, three silver workers and five referents chose not to participate, and thus the exposed and reference groups were reduced to 24 and 22 subjects, respectively. Briefly, the study was conducted as follows. A projected image of light and dark vertical bars was viewed through a translucent diffusion screen which, depending on the subject's response, was moved closer to or away from the stimulus until a final threshold level was achieved. The number of "counts" (steps separating the image from the diffusion medium) was analyzed for the silver workers with and without corneal deposition and for the referents according to various spatial frequency and illumination parameters. The subjects with a large number of "counts" were considered to have a low contrast pattern and high sensitivity. (The study methodology, including a schematic diagram of the instrumentation, is presented in the appendix.)

Pulmonary function
Measurements of age-and height-adjusted forced vital capacity (FVC) and forced evpiratory volume in 1 s (FEVl.o) were calculated on the basis of regression equations derived from a random sample of "healthy" white, nonsmoking males 25 years of age and older (25). To assess the statistical significance of silver exposure and smoking effects (including their interaction), a two-way analysis of variance (ANOVA) was performed.

Chest radiography
Posterior-anterior and lateral chest radiographs were evaluated by a radiologic consultant.

Clinical laboratory studies
For the evaluation of renal function, estimated creatinine clearance, based on the median of three 40-min timed urine collections, and serum creatinine tests were undertaken. The creatinine clearance values were corrected for lean body mass with the use of a body surface area nomogram (26). Regression analysis techniques were used to assess the relationship with estimated career silver exposure . Four liver enzymes (alanine aminotransferase, aspartate aminotransferase, gamma glutamyl transferase, and alkaline phosphatase) were also measured. Other tests included blood urea nitrogen, total cholesterol . complete blood count , and urinaly sis.

Biological monitoring
A study was conducted to estimate the body burden of silver in the recovery workers, including the amount absorbed and excreted in selected biological specimens, and to compare these data with similar information for their matched referents. Samples of blood, urine , feces, and hair were analyzed for total silver by flameless atomic absorption spectroscopy. The methodology and results of a larger study of silver recovery workers and referents have been published earlier (27).

Age and employment history
The average age of the 27 silver workers in 1979 was 46.2 years. Historically, employment in the department has been stable. Nine subjects had at least 30 years of experience. while six had worked for less than a decade. The mean duration of employment was 19.3 (ran ge  years. The referents' average age was 45.6 years, and their average tenure 20.4 years.

Height, weight, and blood pressure
The silver workers weighed an average of approximately 9 kg more than their matched referents (91 versus 82 kg, P < 0.05). Thirty-seven percent and II 0J0 of the recovery employees and referents. respectively. were at least 20 0J0 above their mean age-and heightspecific weight (28). Elevated blood pressures [systolic 140 mm Hg (18.62 kPa) and/or diastolic~90 mm Hg (11.97 kPa)] were reported for approximately one-third of the employees in each group.

Dermatological findings
No cases of generalized argyria were observed. Twenty of the 27 silver workers (74 070) exhibited some degree of internal nasal-septal pigmentation, almost half displaying " trace" amounts; three had "marked" discoloration. In contrast, no nasal staining was observed in the referents. Examinations of the oral mucosa, gums, and face showed no staining in either group of emplo yees.

Facial photograph study
Neither the physicians nor the laymen were able to distinguish the exposed subjects from the referents at rates significantly different from chance. The medical judges selected silver workers 54 070 of the time (P =0.10) compared with 53 070 for the laymen (P =0.16). In addition the mean pairwise differences in the discoloration scores between each exposed subject and his matched referent were not significantly different from zero. The laymen reported substantially more "obvious" pigmentation (8.3 070 of the silver worke rs and 11.5 070 4 of the referents) than the physicians (0.0 and 0.8 070, respectively). In addition, no statistically significant differences between the silver workers and the referents were found when the ANOVA was performed, regardless of the panel composition. The P-values for the physicians alone and for all three panels (clinical physicians, research physicians, and laymen) were 0.63 and 0.87 , respectively.

Electron microscopy
Examination of the granular and basement membrane structures and of the staining pattern of collagen revealed no evidence of silver in any of the biopsy specimens.

Ophthalmologic examination
The most noteworthy ocular manifestation was argyrosis, a staining of the bulbar conjunctiva and its contiguous muco sal surfaces. Compared with the reference group, in which no unusual findings were reported, five of the 24 (21 070) silver workers had grade 1+ or higher conjunctival discoloration, while six (25070) had corneal silver deposition detectable in Descement's membrane (table 2).
Overall, seven (29 070) employees showed some degree of pigmentation in one or both eyes (table 3). Conjunctival staining only was observed in one worker, while two had barely detectable amounts of corneal pigmentation only. The individual with 3 + conjunctival and 2 + corneal deposition had been emplo yed in the recovery department for approximately 32 years, IS of which were in materials handling and washing operations, II in water treatment, 5 in smelting/refining, and I in incinerator operations. His estimated career exposure (see definition given later in the text) was in excess of 2 000 (ug Ag/m')-years. None of the subjects, including the referents, exhibited lens changes or funduscopic abnormalities.   A dose-response analy sis was conducted based on exposure estimates derived from the histori cal job history data , including those from both the questionnaire and personal interview. Work assignments in the silver recovery department (through January 1981) were assessed according to the duration (months) and estimated level (micrograms of silver per cubic meter) of expo sure . The accumulated product, an estimate of potential career silver dose, was expressed in unit s of micrograms of silver per cubic meter-years [(/lg Ag/rrr'j-years]. The study population was divided into three dosage groups of approximatel y equal size, the natural separations in the exposure distribution being taken into consideration. The sum mary statistics were as follows : mean career exposure rate 110 (range 5-240) ug Ag/rrr' and mean career dose 1 900 (range 70-7 000) (/lg Ag/m J)-years.
An evaluation of the argyrosis cases accordin g to estimated career exposure category (table 3) suggested a dose-re sponse relationship (P = 0.06 , chi-square test for linear trend). Further analysis using individual career exposure values indicated that the regression coefficient was significantly different from zero (P = 0.03, chi-square test).
To determine whether silver depos ition was associated with ocular impairment, we divided the exposed subjects into two groups according to the presence (N = 6) or absence (N = 18) of corneal silver deposition ; data for the referents were also anal yzed. On the basis of Student's t-test, no significant differences were found between the two populations of recovery employees or between the silver workers and the referents for any of the standard optometric tests evaluated (refractive error , keratometry, tonometry, and corrected visual acuity).
Visual contrast sensitivity experim ent A comparison of the "counts" with a two-sample t-test indicated that there were no significant differences between the silver workers and the referents at any of the frequency/glare levels (P =0.14 to 0.96) . In general , the contrast sensitivity measurements were higher for nonglare illumination. Contrary to the hypothesis relating lower contrast sensitivity with argyro sis, workers with corneal silver deposition demon-stra ted consistently higher values than those with no ocular pigmentation. However, the differences between the two groups were not statistically significant (P=0.06 to 0.90) .
Respiratory system findings Symptoms. On the questionnaire four silver workers and two referents reported that they were " troubled with a cough almost every day." One recovery employee, a nonsmoker with recurrent respiratory infection s, became asymptomatic following antibiotic therapy. Another, who had smoked a pack of cigarettes daily for 10 years, had symptoms consistent with a diagnosis of chronic bronchitis. The third was a longterm smoker .(31 pack-years), while the fourth had never smoked cigarettes. All had negative rad iographic findings; none had a history of asthma or wheezing.
One referent with a 68 pack -year smoking history indicated that he had had a chronic cough of more than six month' s duration during each of the past two to four years; his chest radiograph was "suggestive of mild chronic bron chitis ." The other , an ex-smoker , had smoked 1.5 packs of cigarettes per da y for 30 years.
Pulmonary function. No significant differenc es in ventilatory performance were observed between the study groups accordin g to smoking category . The mean FVC values of 4.5 and 4.6 1(P = 0.16) for the silver workers and refer ents , respectively, and the mean FEV 1.0 of 3.6 1 for both groups (P = 0.39) were not statistically different. The majority of employees with spirometric abnormalities were chronic tobac co users (30-45 pack-years). Four of the five silver workers and five of the six referents with a decreased FEV 070 [(100 xFEV l.o)/FVC] «75 % ) were smo kers.
Chest radiograph. No abnormalities were observed among the recovery employees , while the one referent with symptoms suggestive of bronchitis was reported to have " mild accentuation of bronchovascular markings in both lung fields." A stable granulomatous lesion was also observed in a silver worker.
Clinical laboratory data Urinary system findings. The mean creatinine clearances of 0.98 and 1.10 ml . S-I . m-2 for the exposed and reference groups, respectively, were not significantly different (P = 0.10, paired t-test) . One silver worker with a history of chronic alcohol abuse had an abnormally low value (0.28 ml . S-1 • m-2 ) . None of the referents had an abnormal creatinine clearance level. An analysis of dose response showed no significant relationship between career silver exposure and creatinine clearance (P = 0.31) . The mean serum creatinine values for the exposed workers and the referents were, respectively, 92 and 87 umol /l (P = 0.14), while the mean blood urea nitrogen level of 6.2 rnrnol/I of the referents was marginally higher than that of the silver workers (5.4 mrnol /I) (P = 0.07). The urinalysis results for protein, glucose, and both red and white blood cells were not statistically different in the two populations.
Other cfinicallaboratory results. None of the differences in liver enzyme levels (alanine aminotransferase, aspartate aminotransferase, gamma glutarnyl transferase, and alkaline phosphatase) between the exposed workers and the referents was of statistical significance. The exposed workers exhibited a marginal decrease in red blood cell count (5.1 versus 4.9.10 6 , P=0.04), and a significant increase in mean corpuscular volume (94.7 versus 90.0 urn", P = 0.02); neither finding was considered clinically important.
Biological monitoring. The mean concentration of silver was 0.010 ug/rnl among the 21 reclamation employees with measurable blood silver levels (2: 0.005 ug/rnl) . In contrast, silver was not detected in the blood of any of the referents. Detectable levels of urinary silver (2: 0.005 ug/g) were observed in only one exposed worker who had normal renal function and in none of the referents. Silver was observed in all fecal samples collected with mean concentrations of 16.8 and 1.5 ug/g for 18 exposed workers and 22 referents, respectively (P<O.OI). The mean silver concentration in the washed hair samples was markedly higher (P<O.OI) in the exposed group (88Ilg/g, N=26) than in the reference group (0.5 ug /g, N = 27). However, this difference may have reflected the potential direct binding of airborne silver particles to the hair rather than its metabolic deposition.

Discussion
This investigation has demonstrated no unusual health patterns in 27 employees with long-term exposure to primarily insoluble silver compounds. On the basis of the findings of reports involving soluble compounds, the following four hypothesized target organs were examined : skin, eyes, lungs, and kidneys.
The principal forms of silver discussed in the literature have been soluble compounds, notably silver nitrate. The health risks associated with exposure to insoluble compounds such as silver halide have been less extensively addressed. Data from non mammalian studies have shown that the toxicity of silver differs substantially depending upon its speciation. LaBlanc et al (29) reported that relatively insoluble compounds, such as silver sulfide and silver thiosulfate complexes, are approximately 15000 times less acutely toxic to fathead minnows than silver in its soluble state (free silver ion). While it is difficult to extrapolate human toxicity potential from investigations of aquatic life, the results demonstrate the need to consider silver speciation in the assessment of human health effects.
Historically, the primary chronic health concern associated with silver has been its skin effects. Electron microscopic studies have determined that silver granules are deposited primarily in the basal lamina of the eccrine sweat glands, in the dermal elastic fibers, and around pilosebaceous structures (10). While a melanin-silver interaction has been suggested (30), silver is more commonly found in nonmelanin-containing tissue (10,31). Such deposition, which causes the slategray appearance characteristic of generalized argyria, has not responded satisfactorily to depigmentary therapeutic agents, including potassium iodide (32), methenamine (33), and sodium thiosulfate (34).
Case reports and clinical studies describing argyria (both local and systemic) and argyrosis have been reported for persons with a history of occupational silver exposure (1-3, 5, 10, 35-37). Occupational epidemiologic studies using matched, unexposed referents have not been reported in the literature.
The earliest documented case of work-related argyria, published in 1872 (38), occurred in a 72-yearold French woman with 50 years of experience as a silver polisher. Case series by Harker & Hunter (35) in 1935 and Hill & Pillsbury (36) in 1939 described localized argyria in II workers (silversmiths, polishers, smelters) exposed primarily to silver metal particles and generalized argyria in 17 individuals (silver nitrate makers and packers, engravers, miners) with industrial exposure to silver compounds (table 4) .
In 1958, Remler (37) reported 31 cases of occupational argyrosis in employees exposed 2-30 years to silver bromide and silver nitrate in the manufacture of photographic and X-ray films. In this study, the cases demonstrated no decrement in visual acuity, field of vision, or color perception. It was hypothesized that argyrosis originates from the external deposition of insoluble silver granules in the conjunctival sac . Once deposited, a chemical reduction to the metal occurs through the combined effects of vitamin C-rich alkaline tears and ultraviolet light.
A 1979 cross-sectional clinical investigation of 30 workers exposed to silver nitrate and silver oxide indicated that six were diagnosed with generalized argyria and 20 with argyrosis (1). A later study by the same author (3) found 10 cases of conjunctival (~2 +) and six cases of corneal (4 +) argyrosis in 27 workers exposed to a variety of silver compounds (nitrate,  (3) stud y sil ver oxide , so lubl e product io n silver c hloride, and main-Ten (37 % j and si x silver cad-l enance (22 % ) emp loy ees , miu m oxide worker respect ively, had co njunc ti val (" 2 + ) and corneal (4 + ) depo sit ion , Inhalati on was the primary route of exposure. Oth er types of exposures were oral (hand -to-mouth conl aminati on , i ngestion via pipelting), ophthalmic (rubbing fi ngers i n eyes), and local (instillation throug h bro ken sk in).
oxide, chloride, etc). The median duration o f silver exposure in both of these studies was slightly more th an five years compared with about 18 years in the current investigation. As noted, limited information is cur rently available concerning the risks of argyria or argyrosis from insoluble silver compounds. In contrast to a clinical study of predominately silver nitr ate workers (I), our investigation of recovery emplo yees exhibited no evidence of systemic argyria as assessed by clinical examinat ion and skin biops y with electro n microscopy and dispersive X-ra y analys is. In addit ion, on the basis o f an assessment of standardized facial photo graphs, layman and physician panels were unable to perceive a difference in either the frequen cy or severity of facial discolorat ion for the silver workers versus that of referents. Th ere was, however, some internal local dis-216 colorati on of the nasal septum, app arently unr elated to the duration of employment and blood silver level. No increased pigmentation of the oral mucosa or gums was observed.
The lower prevalence of conjunctival argyrosis in the present investigation (21 070) in comparison with the prevalence determined in the Rosenman et al studies (I , 3) (67 and 63 %) may have been due , in par t, to differences in silver speciation. In contrast, the corneal argyrosis find ings at a metallurgical refinery (3) were similar (22 %) to tho se reported in the current study (25 %). The proportion was higher (50 %) for th e silver nitrate and silver oxide work ers Despite its frequent clinical presentation, there is general agreement that argyrosis causes no impairment of visual function. Furthermore, an association with ents ralized argyria of occupational origin was at no time a common condition. irs almo st exclu sively in makers of silver nit rat e. Owing to c hanging condln indu st ry it is now fas t disappearing [p 441)." i included 24 cases in addition to the 16 reported by Hark er & Hunter (35) ion of exposure 2 to 30 years nated exposure 22 to 3 10 Ilg /m 3 as metallic silver; restrictive pulmonary tion in two workers although no spirometric measurements given; carbon 'xlde transfer reduced in three em ploy ees te sted ; how ever, data uncorrected making and age; renal and hepatic functi ons normal isure (silver mel al) 39 10 378~g/m ' ; median durati on o f employment appro xi-!Iy five years; creatinine clearance reduced in five subjects (16 % ), ten reported eased night visi on; elec tro ret ino grap hic st udies co nduct ed on seven we re uive, 12 had me asurabl e blood sil ver levels biopsies of four referent s normal; selenium, mercury, tit anium, and iron in cplace environment ge of 8·h tl rne-we lqhted average 40 to 350~gfm'; mean tenure 8.1 years, other snt lal exposures inc l uded cadm ium, formald ehyd e, nitric and hydroch lori de Is, hyd roquinon e, meth anol , and ox ides o f nitrogen; excluding fo ur hypert ens and/or proteinurics, the mean urinary N·acetyl·n·glucosamin idase (NAG) value . signif i cantly high er in th e sil ver workers th an t he comp ari son group of Jsten carbide machine operators; elevated blood sil ver, urinary silver, and mium levels were observed in 92, 96 and 19 % of employees, respectively. nyctalopia has not been demonstrated . One-third of the silver workers studied by Rosenman et al (I) reported decrea sed night vision in their medical history. However , after conducting a series of tests (electroretinography, color-vision screening, visually evoked potentials, etc) to confirm these subjective complaints, Moss et al (2) concluded that, " no electrophysiologic or psychophysiologic deficit could be documented [p 908]." In addition, the result s of contrast sensitivity studies were negative for four individuals tested by the Mt Sinai ophthalmologists, while the present epidemiologic investigation, which evaluated a larger number of subjects, showed no significant difference between the recovery workers and referents .
Silver polishers and finishers exposed to metallic silver, as well as to abrasive pastes and rouges, have been the primary occupational groups represented in the few case reports which have addressed respiratory pathology. In 1945, McLaughlin et al (39) described clinical and radiologic findings for four silver finishers who were chronically exposed (20-40 years) to iron oxide (from rouges), cotton, and silver dusts. Three of the employees had a history of cough accompanied by redblack sputum; no other pulmonary signs or symptoms were reported. Radiographic examination revealed the presence of reticulation in the lung fields, " resembling that found in welders, in haematite miners, and in other workers who inhale iron and iron oxide dust [p 340]." With the exception of one case of emphysema, no physical disability was observed. Silver was not implicated .
Barrie & Harding (40) in 1947 described postmortem examinations of four British silver finishers [including one of the McLaughlin et al (39) cases] who were also exposed to iron oxide and silver dusts . Microscopic and histochemical analyses showed silver impregnation of the alveolar walls and small pulmonary vessels. Three of the four had no history of respiratory disability; one complained of a chronic cough . While all the cases were emphysematous, the authors concluded that these findings were not unusual since, "most Sheffield workmen have well-marked emphysema at the age of 60 [p 228]." No information was available concerning smoking history.
In a 1977 case report, Perrone et al (5) described pulmonary function and chest radiographic results for four Italian silver polishers with long-term (25)(26)(27)(28)(29)(30) years) exposure to silver particles and dusts from abrasive pastes (composition unknown) and fabric and vegetable fibers. Increased bronchovascular markings, reduced lung capacity (restrictive syndrome), and diminished carbon monoxide transfer were noted in two employees, both of whom were smokers (approximately 20 cigarettes per day). It is difficult to interpret the extent to which silver contributed to these findings since the work environment included multiple agents, some of which may have been fibrogenic . In addition , the carbon monoxide transfer testing methodology and comparative values were unreferenced.
There was no evidence of restrictive pulmonary disease in the Rosenman et al (I) study ; obstructive changes were essentially limited to smokers and exsmokers. In addition roentgenographic examination did not show increased bronchovascular markings. However, it should be noted that, while these employees were exposed to relatively high levels of silver nitrate and silver oxide, they were not subjected to the fibers, abrasives, and dusts characteristic of silver polishing .
The negative pulmonary results in the current investigation are consistent with those reported by the Mt Sinai physicians. Cigarette smoking was apparently the most important risk factor associated with both respiratory symptoms and decreased pulmonary function. In addition , the chest radiographs were essentially negative in both the study and comparison groups.
Histochemical analyses of deceased persons with argyria have demonstrated high concentrations of silver in the kidney (41). However, the toxicologic significance of this finding is unclear. Zech et al (4) described a case of nephrotic syndrome with increased blood urea nitrogen and creatinine in a 73-year-old man who had used silver-containing mouthwash for 10years; his estimated cumulative dose of metallic silver was 88 g. Renal biopsy revealed an endarteritis, as well as silver deposits in the basal glomerular membranes. Since no information was available concerning a history of hypertension, diabetes, or chronic medication usage and since a therapeutic response occurred following the administration of an antiinflammatory drug, the etiology of this unusual case of nephrosi s in a person with argyria is uncertain.
Abnormal creatinine clearance values were observed in five of the 30 study subjects examined by the Mt Sinai researchers (I). Hypertension, reported for three of these individuals, could have influenced renal function. The authors indicated that "no other (than blood pressure) etiologic factors were elicited" and that "additional studies will have to be done to further investigate possible kidney damage [p 434]." Compared with the levelsof their unexposed referents, significant-1y lower creatinine clearance values (unrelated to either blood or urine silver concentrations) were observed in the New Jersey metal refinery workers studied by the same principal investigator (3). In addition, the Nacetyl-l3-glucosaminidase (NAG) levels were significantly elevated in the silver-exposed workers (four subjects having extreme values) versus those of the referents. In interpreting this finding, the authors stated that " . . . it is not possible to determine if the observed NAG rises are due to silver deposition or other renal toxins such as cadmium [p 271]." Furthermore, the researchers did not assess the potential confounding effect of aspirin on the NAG levels (42). (This medication may have been used since more than one-half of the study group reported upper and /or lower respiratory symptoms.) The current study, in contrast, found no dose response or significant differences in renal function between the silver recovery workers and their referents, both groups exhibiting similar creatinine clearance, blood urea nitrogen, and urinary protein findings. It should be noted, however, that sensitive measurements for identifying renal dysfunction such as the urinary excretion of 13-2-microglobulin and tubular enzymes were generally unavailable at the time the study was conducted.
The photographs were reviewed by 23 judges assigned to the fo llowing three panels: panel 1: laymen (clerks a nd te chnicians from our laboratory, N = 12); panel 2: research ph ysicians (administrators and toxicologist s from the sa me laboratory, N = 5), and panel 3: clin ical physicians (occupational ph ysicians from two plant medical departments, N = 6).
After the judges read a brief instructional statement, they were given a ca rd de scribing the categories to be used for ranking the individuals according to the extent of their cos metic impairment. Ea ch pair o f portr aits was then examined in sequence. The re viewers were first asked to select the person with facial discoloration and th en to score that individual on a scale o f I to 5; the seco n d ph otogr aph was sco red O. The st ud y subjects were reviewed in pairs rather than individually because of the stro ng association between age and the pr esence of acne, nevi, and other facial blemi sh es. Electron microscopy study Fixed blocks initially immersed in 5 % glutaraldehyde and later -in Dalton's chrome-osmium solution were dehydrated in a graded series of mixtures of ethanol and propylene oxide and embedded in Epon 812®. Thick sections were stained with toluidine blue and examined by brightfield and darkfield microscopy. Thin sections were collected on 200-mesh copper grids. Sections stained with uranyl acetate and lead citrate and unstained sections were examined with a Philip s 201 electron microscope. Selected unstained sections collected on carbon-coated nonmetallic grids were evaluated with a lEOL 100CX electron microscope equipped with a Kevex 7000 energy-dispersive X-ray microanalysis unit.

Contrast sensitivity study
Contrast sensitivity measurements were obtained by the technique known as Wetherill tracking (1). The data were collected by means of a computer-controlled apparatus designed and constructed by a psychophysicist (RTK) in this laboratory (2). A schematic of the test device, which was based on a model developed by Carlson & Heyman (3), is shown in figure I.
A target of light and dark vertical bars in the form of a sinusoidal grating pattern was presented to the subject by rear projection onto a screen from a 35-mm slide. Contrast was altered by the changing of the relative positions of a translucent diffusion screen between the subject and the projected image . Depending upon the subject's response, the diffusing screen was moved a fixed number of steps either closer to or away from the stimulus, the visual contrast pattern thus being altered. In addition to recording total responses, the computer calculated the mean number (and variance) of " counts" (steps separating the stimulus from the diffusing medium) for each threshold level. A large value was associated with a low contrast pattern and high sensitivity.
Prior to the experiment, refractive errors were corrected optimally for each subject using corrective lenses, where appropriate. Following a brief introduction , which included detailed operational instructions, the subject was shown an easily recognizable test pattern. The session was initiated by a switch being pressed which caused two slides, randomly selected as either a blank or a sinusoidal grating, to be displayed sequentially at 0.5-s intervals. After either the first or second was chosen as the grating pattern (by the keying of the appropriate response) , the diffusion screen was positioned to the next stimulus level under program control. This pro cess was repeated until the final threshold level was reached on the basis of a predetermined criterion response probability. For this experiment, in which this value was 0.71 , two consecutive correct responses were required in order to lower the stimulus level (increase the distance between screens), whereas an incorrect response raised it.
Measurements were made for four spatial frequencies (5.9, 7.5, 8.3, and 14.5 cycles per degree of arc subtended by the ocular viewingangle) under both normal illumination and with a bright point light source (incandescent lamp) located approximately 30°in the peripheral visual field. The glare condition could be expected to enhance the effect of light scattering on contrast sensitivity. The eight viewing conditions (four spatial frequencies x two illumination types) were presented to the subject in a partially randomized manner (not more than two of the same kinds of lighting were permitted in succession). Each subject completed the session in approximately 50 min.