Skip directly to search Skip directly to A to Z list Skip directly to site content



Appendix D - Blood Mercury Data

Evaluation of Blood Mercury Data
Table 13 - Blood Mercury Levels of Ciudad Cristiana Residents and Others
Table 14 - Blood Mercury Levels of Ciudad Cristiana Residents before September 1
Table 15 - Blood Mercury Levels of Ciudad Cristiana Residents after September 1

Evaluation of Blood Mercury Data

Testing blood, tissue, or other body fluids from individuals living around a site for possible sitecontaminants can identify whether exposure has occurred and if contaminant concentrations arehigh enough to cause adverse health effects. Many of the over 1000 residents of CiudadCristiana were tested for mercury, with the vast majority of the analyses being for mercury inblood. Urine and hair were tested for mercury; blood from a few individuals was also tested forlead.

Of those various biological analyses, ATSDR has in its records the individual results for severalhundred blood mercury analyses. This data set was the only one suitable for evaluation in thispublic health assessment because there was adequate information on quality control/qualityassurance and a fairly good number of results available.

Review of Other Biological Data for Ciudad Cristiana Residents.

Urine from 795 Ciudad Cristiana residents was tested for mercury in February 1985 by the PuertoRico Department of Health (PRDH)(1). Twenty-one individuals had levels above thenormal/background of 40 µg/l (1). Three of those were above the level, 150 g/l, where symptomsbegin to be observed. Analysis of urine for mercury was reportedly done for 21 CiudadCristiana residents in April 1985 by a consultant for EPA (43). All 21 are above 40 µg/l and 3 ofthose are above 150 µg/l. It is not known whether this was a retesting of the 21 individualsidentified in the PRDH study as having urine mercury levels above 40 µg/l. In a 1988 ATSDRHealth Consultation, 2 of 7 urine samples were above 20 µg/l (44). Additional evaluation ofthose urine mercury data can not be done because of a lack of individual results, and informationon quality control and quality assurance.

There are two reports of testing of mercury in hair from Ciudad Cristiana residents (43,44). Atotal of 47 results were reported with a range of 0.07 - 17.2 ppm.

In March 1985, PRDH obtained blood samples from 993 Ciudad Cristiana residents (45). Theyfound only six individuals with levels higher than 5 micrograms per deciliter (µg/dl). (A deciliteris equal to 100 milliliters.) Those six levels ranged from 10.3 to 50.5 µg/dl. However, ATSDRwas unable to obtain a copy of PRDH's results or any other information about the study such asthe protocol for blood collection and shipment, the laboratory that performed the analyses,analytical method, quality assurance/quality control procedures, etc.

Source of Blood Mercury Data

The blood mercury data evaluated here are from individual clinical reports and lists of individualresults provided to ATSDR by the former residents of Ciudad Cristiana and EPA. Blood wasdrawn at several clinics or physicians in the Humacao area from January - May 1985 withadditional samples taken as late as 1993. Collection of samples was done on a voluntary basis.

Nearly all of the analyses were performed by Roche Biomedical Laboratories. It is assumed inthis evaluation that the reported results were for total mercury because there was no indication onthe lab report forms that anything other than the standard analysis was done. Roche indicatedthat they analyze for total mercury by atomic absorption (AA) on digested, extracted wholeblood, unless otherwise noted (43). Two literature reviews indicated that this is and has been thestandard procedure for analyzing mercury in blood for 30 years (16,29). No information on thequality control/quality assurance protocols used by Roche and the other labs could be obtained,though attempts were made.

Information on residence, age, and sex was identified from the clinical reports or from lists ofresults provided by the former residents or EPA. The methods for collecting, handling, andshipping the blood samples were provided to ATSDR by the clinic where most of the bloodswere drawn.

Description of Blood Mercury Data

Three hundred twenty-eight individual results were identified through review of the datasubmitted to ATSDR by EPA and the former Ciudad Cristiana residents. Most were submittedto ATSDR in 1986, but the former residents continue to submit results of analyses. The name,blood mercury results, date of collection, and, if available, name of laboratory, age, sex, andaddress at Ciudad Cristiana were entered on a computer using EPI INFO software. The resultinglist was reviewed to delete duplicate entries. This was done through comparison of names,collection date, analytical results, and demographic information. Eight duplicates wereidentified, reducing the number of individual results evaluated to 320. Of the 320, 266 wereconfirmed by individual laboratory reports. Several individuals had their blood analyzed formercury more than once. Each of those different analyses are included in this evaluation. UsingEPI INFO, a descriptive analysis (mean, range, and standard deviation) by age and exposuregroup was made of the data.

The individual results were divided into two groups - those whose blood mercury levelsrepresented mercury exposure as a resident, and those whose blood levels could not be due toexposure while a Ciudad Cristiana. The first group were those individuals whose blood wascollected before September 1, 1985 and the second group after September 1. The evacuation ofCiudad Cristiana was completed in early May 1985. Mercury has a half-life in blood of 60-70days. Therefore, at least 3/4of any mercury in the blood when a resident left in May would havebeen eliminated from the blood by September.

The results of ATSDR's evaluation of the blood mercury data are described in Tables 13 - 15,which follow this evaluation. For those Ciudad Cristiana residents whose blood was collectedbefore September 1, 1985, the blood mercury level ranged from 0 - 16.8 µg/dl with a mean of 3.9µg/dl. The blood mercury levels for those residents whose blood was collected after September1, 1985 varied from 0 - 18.5 µg/dl with a mean of 2.7 µg/dl. Blood was collected beforeSeptember 1, 1985 from three individuals who worked at or near Ciudad Cristiana, but who werenot residents. Their levels were 5.7, 8.2, and 10.3 µg/dl.

Limitations of Blood Mercury Data

Limitations of the blood mercury data evaluated include the short half-life of mercury in humanblood, the self-selection of participants, and the less than complete information on quality controland quality assurance. Mercury in human blood is an excellent indicator of ongoing exposuresand the risk of health effects, but, because of the 60 - 70 day half-life, blood mercury levels donot identify past exposures. For example, mercury levels in blood from a former CiudadCristiana resident taken in May 1986 reveal nothing about the mercury exposures that a personhad while living in Ciudad Cristiana. Any mercury in the blood in 1986 would have beeneliminated after a year. Any mercury present in the blood in May 1986 would have come fromexposures in the past 3 or 4 months. Therefore, only the blood mercury levels obtained beforeSeptember 1, 1985 can measure the possible impact of exposure to mercury while a resident ofCiudad Cristiana.

The self-selection (voluntary participation) of Ciudad Cristiana residents for blood collection andanalysis limits the blood mercury data through the possible inclusion or exclusion of residentswhose mercury levels are much higher or lower than the average resident. Only by testing everyresident or a random sample of the residents can the true average blood mercury level beidentified.

A related limitation is the great difference in the results observed in this data set of individualresults from 320 residents where 28% of those tested had blood mercury levels greater than 5.0µg/dl, and those for the PRDH study where less than 1% of the 993 participants had levelsgreater than 5.0 µg/dl. This difference between these two data sets which were collected at aboutthe same time, brings into question the validity of both data sets.

Another limitation is the apparent lack of a systematic follow-up of those tested. Typically, anyindividual found to have a level above the laboratory reference level should be retested, and, ifthe level is confirmed, possible exposure sources investigated (26). There was no indication ofany follow-up of those whose data are evaluated here or of those in the PRDH study. Properfollow-up could have provided specific information on the various possible exposure sources.

The main gap for quality control/quality assurance of these data is the lack of information onlaboratory quality methods. Several unsuccessful attempts were made by ATSDR staff to obtainthis information.

Identification of "Normal" Blood Mercury Levels

Recent literature reviews indicate that most populations have low levels of mercury in their blood(16,29,46). Those low or "normal/background" blood mercury levels are due to a variety ofsources other than work. Those sources will be evaluated in detail in a subsequent subsection.

A review of about 80 recent studies of mercury in human blood indicates that normal levels varybased on amount of fish consumed (46). The studies used in this review were of different racial,cultural, or ethnic populations throughout the world. For populations where no fish were eaten,the mean blood mercury level was 0.2 µg/dl with 90% of the values between 0 - 0.4 µg/dl. Where less than two meals of 200 grams of fish each meal are consumed a week, the mean bloodmercury level was 0.48 µg/dl with 90% of the values from 0.24 - 0.72 µg/dl. For populationswho eat 2 - 4 fish meals of 200 grams a week or more than 4 meals, the means were 0.84 and4.44 µg/dl with 90% of the values from 0.26 - 1.42 and 0.61 - 8.27 µg/dl, respectively. Forpopulations where fish consumption was unknown, the mean blood mercury level was 0.58 µg/dlwith 90% of the values from 0.12 - 1.04 µg/dl. Another literature review derived a "normal"blood mercury level of 0.8 µg/dl based on a combination of all the available non-occupationalblood mercury data (16). This value represents a blending of different fish consumption rates.

Many of the laboratory reports on blood mercury levels from Roche identified a reference values. All those reports from 1984 and 1985 have a reference value of 2.8 µg/dl, while those from 1986- 1992 did not identify a reference value. The one result from 1993 listed a reference value of 5.0µg/l, which is 0.5 µg/dl. Typically, a reference value is what a laboratory considers normallevels, based on the sum of all analyses done for that substance by that laboratory. Severalunsuccessful attempts were made to obtain an explanation from Roche for the nearly 6-folddifference between the reference value listed in 1984-85 and the one for 1993.

Identification of "Safe" Blood Mercury Levels

Attempts have been made to identify a "safe" blood mercury level (i.e., the blood mercury levelbelow which adverse health effects are unlikely to occur) (16,22,29). The "safe" blood mercurylevels vary according to whether exposure is mostly to elemental mercury vapor or to organicmercury. For this evaluation, the "safe" blood mercury levels for organic mercury will be used. Because mercury in the environment is mostly converted to organic mercury, residents at CiudadCristiana were exposed mostly to organic mercury (16).

For exposure to organic mercury, adverse health effects begin to be observed in adults at 20.0µg/dl in a study where the route of exposure was ingestion of grain contaminated with organicmercury (29). In the same study, adverse effects on the fetus, which is most sensitive to theeffects of organic mercury, began to be seen at a blood mercury level in the mother of 4.0µg/dl.(1) However, health effects might have occurred in some of the unborn babies of motherswhose blood mercury levels were above 4.0 µg/dl. Twenty-seven of the 64 women (42%) ofchild-bearing age tested had levels above 4.0 µg/dl. The health effects that might be observedat mother's blood mercury levels of 4 - 20 µg/dl are delayed motor development and delayedlanguage development (29-33). This means that children exposed to these levels while in theirmother's womb may have walked later than the normal maximum of 19 months, or talked laterthan the normal maximum of 26 months (33). A study done of Iraqi children indicates that about25% of the children with maternal blood mercury levels of 4 - 20 µg/dl would have walked ortalked later; the rest with this range of maternal blood levels had no adverse health effects. It isnot known whether those developmental delays would be permanent or temporary. At maternallevels of 4 - 20 µg/dl, more severe effects such as mental retardation, deafness, blindness,microcephaly (small brain), and cerebral palsy are not observed (30,31).

Interpretation of Blood Mercury Results from Ciudad Cristiana Residents

The mean blood mercury was 3.9 µg/dl for Ciudad Cristiana residents while they lived there(Table 14 which follows this evaluation). Whether this mean level is "normal" or "abnormal" isdependent on the average fish consumption for Ciudad Cristiana residents. A review of theliterature indicates that the mean blood mercury level was 4.44 µg/dl for individuals consumingfive or more meals (200 grams of fish/meal) a week (46). As will be discussed, the averagefishermen in the Frontera Creek area consume the equivalent of 10 fish meals a week, so"normal" levels for Frontera Creek area residents would be expected to be higher. Also, as willdiscussed in the next section, there are a number of other possible sources of exposure tomercury.

None of the 320 samples evaluated here were above 20 µg/dl, the level where health effectsbegin to be observed. About 42% (27/64) of the women of child-bearing age (16 - 45 years old)had blood mercury levels above 4.0 µg/dl. Blood mercury levels above 4 µg/dl in some pregnantwomen could have resulted in delayed motor development and delayed language development intheir children exposed while in the womb, but not more severe effects.

Sources of Exposure to Mercury

The average blood mercury levels of Ciudad Cristiana residents are "normal", but where did themercury in their blood come? Occupation and eating contaminated fish are the major sources ofexposure to mercury, but many other sources have been identified (16,29). These various sourceswill be discussed in this subsection.

Occupational Sources - Work place exposures to mercury include chlorine-alkali plants,cinnabar mining, the manufacture and use of thermometers and other instruments that usemercury, use of mercury compounds as disinfectants in industrial processes, and dentistry(16,29). The exposure would be primarily to elemental mercury vapor in all those occupationsexcept using mercury compounds as disinfectants. The work place sources possible for theCiudad Cristiana residents are the use of mercury compounds as disinfectants in industrialprocesses, manufacture of thermometers, and dentistry. During the time Ciudad Cristiana wasoccupied, mercury compounds were used at the Technicon facility across the highway from thesubdivision and thermometers were manufactured at the Becton Dickinson plant in Juncos,which is 10-15 miles away (50). Dentists and dental hygienists receive low level exposures tomercury vapor when they use dental amalgams (fillings) that contain mercury (16). From thereview of the data provided by the former residents, at least one dental hygienist lived in CiudadCristiana.

Fish Consumption - The major non-occupational source of mercury exposure is consumption ofcontaminated fish (16). Blood mercury levels as high as 20.0 µg/dl have been identified infish-eating individuals (46). "Normal" blood levels in a population vary according to the numberof fish meals eaten a week. Adverse health effects have been identified in populations that eatcontaminated fish (48,49).

Some Ciudad Cristiana residents could have been exposed to mercury through the ingestion ofFrontera Creek area fish based on investigations of fishing practices and fish contaminant levels(4,51,52). While mercury levels in Frontera Creek area fish were not higher than other areas ofPuerto Rico, there were some individuals who ingested mercury at a greater rate than the healthguideline (the EPA reference dose [Rfd] for methyl mercury) of 0.0003 mg/kg/day. Most (72%)of the individuals queried in the 1988 fishing practices survey lived within two miles of FronteraCreek and the Frontera Lagoons (51). The average consumption of area fish and crabs was 284grams or over ½ pound a day (51). Using 200 grams for an average meal, this consumption rateis the equivalent of 10 meals a week. There is a need to inform and advise those eating fish atthis high rate that there is a possibility of health consequences, especially for pregnant women.

About 10% of Ciudad Cristiana residents caught and ate fish and crabs from the Frontera Creekarea according to a survey done by the ex-residents. Therefore, Ciudad Cristiana residents couldhave consumed greater than average levels of mercury, if their daily consumption rate wassimilar to that for the other nearby residents of the area surveyed in 1988.

Other Exposure Sources - There are several other non-occupational sources that may havecontributed to the mercury identified in Ciudad Cristiana residents' blood. Those includemercury in air, latex house paint, dental fillings, certain skin creams and soaps, and residentialsurface soil; accidental release of mercury; and a religious practice called Santería (16,53).

Exposure to air-borne mercury in Ciudad Cristiana may have occurred if the exposure situationidentified during the remedial investigation in 1988 was also occurring when the sub-divisionwas occupied. The mercury in the air was discussed in detail in the Pathways Analyses andToxicological Evaluation sections.

Phenylmercuric acetate, an organic mercury compound, was used as a fungicide in latex paintuntil recently (54). Levels were to be no greater than 300 ppm in interior paint and 2000 ppm inexterior paint. At least one case of childhood mercury poisoning is known from inappropriatelyformulated interior latex paint where there was over 900 ppm of phenylmercuric acetate. Theinside of the homes at Ciudad Cristiana were painted; ATSDR assumes that this paint containedmercury to retard growth of mold and fungus due to the high humidity in Puerto Rico.

There is some research indicating that individuals with dental fillings (amalgams) made withmercury demonstrate slightly higher mercury levels than people without fillings (16). It isgenerally agreed that those higher levels are not related to adverse health effects.

As discussed in the Toxicological Evaluation section, the mercury in the Ciudad Cristianasurface soil does not represent a health hazard. However, the mercury in the soil could be asource for low level exposure, especially for small children who typically ingest more soil thanadults. Those low level exposures might have resulted in detectable levels of mercury in blood.

Some skin-lightening creams and soaps contain mercury and are known to result in highermercury levels in those individuals using them (16). It is not known whether anyone in CiudadCristiana used such creams and soaps.

There are a number of reports indicating that the spilling or misuse of mercury-containingproducts can be another possible source of mercury exposure (16). Two recent examples are thepoisoning of a 15- and 11-year old in Ohio through the spilling of elemental mercury on a carpet,and the death of four adults in Michigan by the attempted smelting of silver out ofmercury-containing dental fillings (55,56). The poisoning of children through the misuse ofmercury has occurred in Puerto Rico.(2)

Another possible exposure source is the use of mercury in a religion called Santería that ispracticed by some ethnic groups in the Americas (53,57,58). In this religion, mercury is used inseveral rituals. Mercury can be carried by individuals in small vials, mixed with soap and waterand used as a cleaning agent, mixed with perfumes, placed in candles, etc. Those practices couldlead to exposure or even mercury poisoning. However, many of the residents of CiudadCristiana were members of a Pentecostal group, and thus would probably not participate inSantería.

Table 13.

in Years
0 - 2.8 µg/dl2.9 - 5 µg/dl5.1 - 10 µg/dl10.1 - 20 µg/dlOver 20 µg/dlMeanTotal by AgeGroup
0 - 535 (48.6%)17 (23.6%)11 (15.3%)9 (12.5%)04.072
6 - 1014 (41.2%)7 (20.6%)10 (29.4%)3 (8.8%)04.434
11 - 1715 (44.1%)9 (26.5%)4 (11.8%)6 (17.6%)04.534
18 - 6558 (45.0%)29 (22.5%)38 (29.5%)4 (3.1%) 03.8129
Over 652 (20.0%)6 (60.0%)02 (20.0%)05.710
AgeUnknown28 (68.2%)8 (19.5%)4 (9.8%)1 (2.4%)02.641
Total byLevel152 (47.5%) 76 (23.8%)67 (20.9%)25 (7.8%)03.9320

* Three hundred seventeen of the 320 samples were obtained from former residents of CiudadCristiana. Three were from individuals who lived elsewhere, but who worked at or near CiudadCristiana.

See the evaluation section of this appendix for an explanation of the source of those data.

Table 14.

in Years
0 - 2.8 µg/dl2.9 - 5 µg/dl5.1 - 10 µg/dl10.1 - 20 µg/dlOver 20 µg/dlMeanTotal by AgeGroup
0 - 534 (49.3%)16 (23.2%)10 (14.5%)9 (13.0%)04.169
6 - 1013 (40.6%)7 (21.9%)10 (31.5%)2 (6.3%)04.332
11 - 1711 (36.7%)9 (30.0%)4 (13.3%)6 (20.0%)05.030
18 - 6553 (44.2%)29 (24.2%)34 (28.3%)4 (3.3%) 03.8120
Over 651 (14.3%)5 (71.4%)01 (14.3%)04.97
AgeUnknown21 (65.6%)7 (21.9%)2 (6.3%)2 (6.3%)02.732
Total byLevel133 (45.9%) 73 (26.1%)60 (20.7%)24 (8.3%)03.9290

See the evaluation section of this appendix for an explanation of the source of those data.

Table 15.

in Years
0 - 2.8 µg/L2.9 - 5 µg/L5.1 - 10 µg/L10.1 - 20 µg/LOver 20 µg/LMeanTotal by AgeGroup
0 - 51 (33.3%)1 (33.3%)1 (33.3%)003.33
6 - 101 (100.0%)00000.641
11 - 174 (100%)00000.74
18 - 655 (62.5%)03 (37.5%)002.88
Over 651 (33.3%)1 (33.3%)01 (33.3%)07.63
AgeUnknown6 (75.0%)1 (12.5%)1 (12.5%)001.78
Total byLevel18 (66.7%)3 (11.1%)5 (18.5%)1 (3.7%)02.727

See the evaluation section of this appendix for an explanation of the source of those data.

Appendix E - Summary of Previous Environmental Sampling for 1978 - 1985

Table 16 - Summary of Previous Soil Sampling Efforts for Mercury
Table 17 - Summary of Previous Sediment Sampling Efforts for Mercury
Table 18 - Summary of Historical Data for Frontera Creek Sediments
Table 19 - Summary of Previous Surface Water Sampling Efforts for Mercury
Table 20 - Summary of Previous Biota Sampling Efforts for Mercury
Table 21 - Summary of Previous Surface Water Sampling for HSL Parameters
Table 22 - Summary of Previous Sediment Sampling for HSL Parameters
Table 23 - Summary of Previous Biota Sampling Efforts for HSL Parameters
Table 24 - Summary of Mercury Usage by Site Industries
Table 25 - Summary of Usage of Potentially Hazardous Chemicals

Source - Remedial Investigation, Volume 2 and 4 (4)

The following tables were not available in electronic format for conversion to HTML at the timeof preparation of this document. To obtain a hard copy of the document, please contact:

Agency for Toxic Substances and Disease Registry
Division of Health Assessment and Consultation
Attn: Chief, Program Evaluation, Records, and Information Services
Branch, E-56
1600 Clifton Road NE, Atlanta, Georgia 30333

Appendix F - Figures 1 and 2

Figure 1 - area map

Figure 2 - specific site area

Source for Figures - Remedial Investigation (4)

Figure 1. Frontera Creek Site Location Map

Figure 2. Specific Site Area for Frontera Creek NPL Site

Appendix G - Public Comments



The Frontera Creek Public Health Assessment was available for public review and commentfrom August 12, 1994 through October 12, 1994. The Public Comment Period was announced inthe El Nuevo Diá and San Juan Star, which are island-wide newspapers, the El Oriental, aHumacao-based weekly, and on WALO, a Humacao radio station. Copies of the public healthassessment were made available for review at the home of Jose Sepulveda, the leader of theCiudad Cristiana Ex-Residents Group, the Antonio Roig Public Library in Humacao, theHumacao College Library, and the EPA Caribbean Field Office in Santurce. In addition, thepublic health assessment was sent to eight persons or organizations. Comments were receivedfrom Applied Geotechnical and Environmental Service Corporation (AGES), the environmentalconsultant to Squibb Manufacturing, Puerto Rico Environmental Quality Board, and the PuertoRico Department of Health.

Comments and responses are summarized below. The comment letters can be requested from ATSDR through the Freedom of Information Act.
COMMENT:We have reviewed ATSDR's Public Health Assessment of Frontera Creek inHumacao, Puerto Rico and feel that the conclusions arrived at are based on athorough and correct analysis of the data available.

RESPONSE:Thanks for your concurrence with the public health assessment'sconclusions.

COMMENT:For us it is obvious that if there had been gross contamination of CiudadCristiana by mercury at least some of the surface or subsurface samplesshould have shown levels above background and they did not show in thesystematic sampling done under the supervision of EPA in the remedialinvestigation conducted in Ciudad Cristiana in 1988.

RESPONSE:ATSDR appreciates your opinion.

COMMENT:We coincide with ATSDR that some of the symptoms and illness reportedby ex-residents of Ciudad Cristiana may have been caused by methylenechloride concentrations in the air; also, it is clear for us that since theparticipants were self-selected there was a major limitation with the bloodmercury data reviewed by ATSDR, since this could result in a weighing ofthe results either to the high or low side. Nevertheless, none of theindividual blood mercury-levels was above the 20 mg/dl level where healtheffects appear to begin to be observed.

RESPONSE:Thanks for your opinion.

COMMENT:The Department of Health of Puerto Rico understands the legal responsibilityof ATSDR to plan public health actions needed at hazardous waste sites butrespectfully it seems paradoxical to us that need exists to issue a fishconsumption advisory on mercury "for those areas of Puerto Rico where acombination of background mercury levels and high fish consumption ratescould result is a hazard to health," when the public health assessment statesthat the soil mercury levels at Ciudad Cristiana are within background(natural) ranges for the area and not due to contamination (pages 37 and 38)and that the mercury levels in Frontera Creek area fish are typical for PuertoRico (page 35). If the need really exists we are willing to receive ATSDRtechnical advice on deriving and issuing the fish consumption advisory, butas stated the mere high consumption of fish in Puerto Rico, as elsewhere,would not justify a fish consumption advisory.

RESPONSE:The Public Health Action Plan has been revised to reflect this.

COMMENT: The Puerto Rico Environmental Quality Board (PREQB) received thereferenced document for evaluation and comments. This report wasperformed to determine whether site-related adverse health effects have, are,or may occur; and to recommend actions to reduce, prevent, or furtheridentify the possibility of adverse health effects. After evaluating thereferenced document, PREQB approves the proceedings presented in thisstudy and concurs with the preliminary results demonstrated hereby. PREQBrecommends to continue additional monitoring of the air in the FronteraCreek area for methylene chloride.

RESPONSE:ATSDR appreciates your opinion.

COMMENT:ATSDR did not perform a critical assessment of the data it utilized, forexample: how limited sampling periods, sample locations or environmentalfactors could affect results. The lack of an evaluation of the data utilized wasparticularly critical in the case of the TRI report because during the earlyreporting years data collection was unrefined.

RESPONSE:ATSDR did perform critical assessments of the data utilized. TheEnvironmental Contamination and Other Hazards section containsdiscussions of the TRI data (p. 9), environmental sampling (p. 10), andlaboratory quality assurance and control (p. 17). There are also numerousmentions of the limitations of data used throughout the public healthassessment.

The discussion of the TRI data specifically mentions the problem ofoverestimation of release amounts. EPA has concluded that this wasespecially a problem in the first couple of years of reporting. However, thisproblem does not appear to affect the TRI data used in this document. Theportion of air releases that were non-point source was over 50% for fouryears (1987 - 1990), then decreased to about 10% in 1991 and 1992. It isvery unlikely that this dramatic change could be due to unrefined datacollection.

COMMENT:The ATSDR Frontera Creek PHA acknowledges that the human bodyeliminates methylene chloride, once introduced. However, it does notaddress how this fact could modify their conclusions. In performing theiranalysis, ATSDR, assumes constant human exposure to the ambient airmethylene chloride present, albeit at two (2) substantially differentconcentrations levels, one (1) high and one (1) low. Consequently, thisassumption results in a gross simplification of the complex interactionbetween transient ambient air methylene chloride concentrations and humanhealth.

RESPONSE:The public health assessment has been revised to clarify this issue. If the airdata for 1992 and 1993-94 do approximately represent the methylenechloride levels in 1983 - 1985, there were periods where continuousexposure did occur, thus creating the conditions where health effects may bepossible. This has been made clear in the revised public health assessment.

COMMENT:ATSDR assumes that methylene chloride concentration levels equal to orhigher than the 1992 levels were present during the entire time that CiudadCristiana was occupied, 1979 to 1985. Up until 1983, methylene chloridewas used in only one manufacturing process at the SMI manufacturing site. Methylene chloride usage was gradually increased until it peaked in 1992. The extrapolation of 1979 to 1985 methylene chloride levels based on the1987 to 1992 data is erroneous. Methylene chloride utilization informationwas available, but apparently were not used by ATSDR in developing itsassumptions.

RESPONSE:ATSDR was not aware of methylene chloride utilization information for theperiod prior to 1987 or it would have been used in its evaluation. The newinformation provided by AGES has lead to revisions in the PathwaysAnalyses and Toxicological Evaluation sections of the public healthassessment. It is now concluded that there was no significant exposure tomethylene chloride from 1979 - 1982.

Incidentally, the TRI data do not support AGES's observation that methylenechloride levels increased gradually from 1983 - 1992. The total air releasesof methylene chloride in 1987 were 103,300 pounds, which more thandoubled to 226,140 pounds in 1988. The total air release bySquibb/Humacao increased slightly to 233,520 pounds in 1989, thendecreased about 20% in 1990 to 188,900 pounds. There was a 37%decrease in 1991 to 119,500 pounds, then air releases leveled out to 118,800pounds in 1992.

COMMENT:There is no correlation between the list of symptoms reported by CiudadCristiana residents and the methylene chloride-related symptomshypothesized by ATSDR.

RESPONSE:The health effects possible for resident workers were subclinical changes inthe liver and kidney and reversible decreases in vision and hearing. Asindicated on pages 33 and 34, both vision and hearing problems werereported. Subclinical changes in the kidney and liver would not be reportedbecause subclinical, by definition, means without symptoms. As indicated inthe public health assessment, most of the self-reported health outcomes,including vision and hearing problems, can be associated with a widevariety of chemical and biological agents.

COMMENT:During the 1990 ambient air monitoring (daytime samples only) presented inthe 1990 RI, acetone was never found at Ciudad Cristiana. That acetonemight have been found at Ciudad Cristiana during evening and night times isspeculation on the part of ATSDR. No evidence supporting the presence ofacetone at Ciudad Cristiana has ever been presented.

RESPONSE:Reported releases of acetone for 1987 - 1992 were over 100,000 pounds ayear. The lack of detection of acetone at the two Ciudad Cristianamonitoring stations during 1989 sampling (there was no air monitoring in1990) is because samples were taken only during the day and becauseCiudad Cristiana is upwind from Squibb during the day under normalmeteorological conditions. If the 1992 and 1993-1994 sampling programshad included acetone as a analyte, it is very likely that it would have beendetected at low concentrations. This is based on that the amounts of acetonereported in TRI are similar to the amounts of methylene chloride released. Since low levels of methylene chloride were detected, acetone would alsohave been present.

COMMENT:The assertion that methylene chloride was being released into the atmospherein mass quantities beginning in 1970 (the SMI manufacturing site start-update) is unsupported. Methylene chloride usage was not introduced at theSMI manufacturing site until 1975, and it was used in very limited quantities(compared to the peak usage year of 1992) until 1983.

RESPONSE:Thanks for the information on when methylene chloride use began at Squibb. The public health assessment has been revised accordingly. Incidentally, thepeak release of methylene chloride was in 1989, according to TRI.

COMMENT:Methylene chloride is not a known carcinogen as was implied in thestatement on page 43; it is a class 2B probable carcinogen as stated on page31. The animal study results used to produce this classification containsubstantial amounts of conflicting data. No human data proving thecarcinogenicity of methylene chloride exists.

RESPONSE:ATSDR identifies a chemical as a carcinogen whenever it is designated assuch by the National Toxicology Program (NTP), International Agency forResearch on Cancer (IARC), U.S. Environmental Protection Agency (EPA),or United States Occupational Safety and Health Administration (OSHA). EPA designates methylene chloride as a B2 carcinogen based on sufficientanimal data but little or no human data. Please consult ATSDR's PublicHealth Guidance Manual for additional information on how ATSDRidentifies a chemical as a carcinogen.

COMMENT:In Table 8, Completed Exposure Pathways, on page 62, Squibb is listed as asource of mercury in ambient air. It is our belief that this is a typographicalerror that should be corrected.

RESPONSE:Thanks for bringing this to our attention. Table 8 has been revisedaccordingly.

COMMENT:Consistent with its policy of being environmentally proactive, Bristol-MyersSquibb Co. is continuing to explore opportunities to modify its processes,replace less desirable solvents and upgrade its emission control equipment atSMI manufacturing site. This approach is consistent with the company'sparticipation in EPA's voluntary Toxic Reduction Program or 33/50Program.

RESPONSE:Squibb's efforts to reduce emissions are mentioned in the public healthassessment.

1. This number was derived by converting the 10 µg/g level in maternal hair identified inClarkson to a blood mercury level using a hair:blood conversion factor of 250:1 (47).

2. Personal communication between Dr. Mark Rodriguez, ATSDR and Dr. Kenneth Dominguez,CDC.

Table of Contents The U.S. Government's Official Web PortalDepartment of Health and Human Services
Agency for Toxic Substances and Disease Registry, 4770 Buford Hwy NE, Atlanta, GA 30341
Contact CDC: 800-232-4636 / TTY: 888-232-6348

A-Z Index

  1. A
  2. B
  3. C
  4. D
  5. E
  6. F
  7. G
  8. H
  9. I
  10. J
  11. K
  12. L
  13. M
  14. N
  15. O
  16. P
  17. Q
  18. R
  19. S
  20. T
  21. U
  22. V
  23. W
  24. X
  25. Y
  26. Z
  27. #