EPA 560/6-80-001
OCTOBER 19 8 0
HEALTH ASSESSMENT OF
OCCUPATIONAL EXPOSURE
TO POL Y BROMINATED BIPHENYL (PBB)
AND POLYBROMINATED BIPHENYLOXIDE (PBBO)
ANITA K. BAHN*
ORA BIALIK0
JACQUELINE OLER°
LORNE HOUTEN0
EMANUEL LANDAU+
^DECEASED
°UNIVERSITY OF PENNSYLVANIA
SCHOOL OF MEDICINE
PHILADELPHIA, PENNSYLVANIA
American public health association
WASHINGTON, D.C.
PROJECT OFFICER
JANE KELLER
OFFICE OF PESTICIDES AND
TOXIC SUBSTANCES
WASHINGTON, D.C.
-------
Sample C. Technical Report Da:a Shoot, EPA Form 2220-
TECHNiCAS. REPORT DATA
(Pfc.7 se re id on >eht t.nxne / cforc
1. AfcJ-ORT NO.
Jj6Qifi=
4. Ti t LE ANO SUOTITLt
health Assessment of Occupational Exposure to
Poly branina ted Biphenyl (PBB) and Polybrorrir.ated
B.iphenyloxide (PBBO)
3. RE
F?31 -1 5<>*75
5. ntf»oHr date
October 1980
6. PEAFCKWING CHGANIZATION COD6
7. AUTHORS) ^
Anita K. Bahn ora Bialik0 Jacqueline 01er°
Landau+
8. PERFORMING ORGANIZATION REPORT NO.
S. PtnFOBMIIMG ORGANIZATION NAME ANO ADO^gSS
American Public Health Association
1015 Fifteenth Street, N.W.
Washington, D.C. 20005
10. PROGRAM ELCMtNT NO.
11. CQNTRACT/GflANT NO.
68-01-3859
12. SPONSORING AGENCY NAME ANO ADDRESS
Office of Pesticides and Toxic Substances
U.S. Environmental Protection Agency
401 "M" Street, S.W.
Washington, D.C. 20460
13. TYPE OP REPORT ANO PERIOD COVERED
Final Retx>rt
14. SPONSORING AGENCY code
15. SU?PLEMENTAHY NOTES
Deceased* University of Pennsylvania0
American Public i tea 1th Association4"
16. ASS fHAGT
In August, 1978, a comprehensive medical evaluation was conducted on 42
out of 96 workers frcrn a plant that manufactured decabror-obiphenyl and
decabromobiphenyl oxide and on 96 control workers fran neighborhood .industry
not' involving PBB use'.' "The exposed workers•'were orployed in a -plant-whidi •• ' .
manufactured only brominated products', where' 96 regular enployee3 had been enployed
since January 1973. PB3 serum level were significantly higher anong the exposed
group than among the non-exposed group. An unexpectedly high prevalence of primary
hypothyroidism was found among the PBB workers. Their sural sensory and
peroneal motor velocities were significantly reduced when contrasted with
controls, however, no significant denra no logical, neurological or other adverse
health effects were clinically daronstrated.
17.
ICCiV WORDS AND DOCUMENT A ^ A L * S1 5
3. DESCRIPTORS
!>.incNTiFiear./crsN ended terms
C. COSAT! fidd/GrOup
Polybroroinated biphenyls (PBB.i)
Polybrominated biphenyl oxides (PBBCs)
Occupational Ex|xjsure
hy^xithroidism
Neurologic Study
Occupational
Thyroid Function
Serum level s
Electrophysiologic Study
Univariate Analysis
?a. DifiTHiBUTlDM STATE Ml: MT
Unlimited
19. S6CURI r Y CLASS f 1 ha firp,>rt/
Unclassified
21. NO. Ok PAGtS
fi?
20. SECUHITY CLASS tThis psjtj
22. PRICE
EPA Farm 2220-1 (9-73)
-------
ACKNOWLEGEMENTS
The authors wish to thank all the workers who constituted the study
peculation and without whose participation the study could not have been
conducted. Our thanks also to the manaqement officials of the company, who
chose to be anonymous, for their complete cooperation. The local unions of
the International Brotherhood of Electrical Workers and of the National Steel-
workers Union were of significant help.
Our special thanks go to the many peoole who helped us in carrvinq out
this study.
First, there are all the staff who Darticioated in the field study:
Peter Gann, M.D., Robert Lovelace, M.D. , Aaron Miller, M.D., Dominic Tarn,
M.D., Theodora Tsongas, Ph.D., Betty Hughley, B.A., Mary Levitsky, R.M.,
Lorette Peterson, B.Sc.; and the interviewers and administrative staff:
Thomas Burke, Patrick Hessel, M.S., Carol Jaadeo, Robert Knorr, M.P.H.,
Margaret Laucihlin, Daniel Laurent, Jonathan Mann, D.D.S., Alma Price, Antolina
Mendez, Martha Hanson, Elizabeth Houten, Nancy Mayerik, Larry Fink, Ruth
Murphy, Florence Cohen, Joan Brunswick, Eileen Lynch and Maryanne Rosowski;
ir. the analysis: Peter J. Snyder, M.D., Marc: Filsteiri, M.D., James Hills,
M.D., Lena Hollmann, K.P.H., and Robert D. tJtiqer, M.D.; and in the 1abora-
troies: Etio Pellizzari, M.D., Ruth Zweidinqcr, Ph.D., Leslie Shaw, Ph.D.,
and Carl Aloer, Ph.D.
Last we would also like to thank Lenora Barnes, M.P.H., and the entire
staff at the American Public Health Association; Josenh Seifter, M.D.,
Frederick J. DiCarlo, Ph.D., Charles Poole, M.P.H., of the U.S. Environmental
Protection Aqencv; and David Linskv of a state unit of the U.S. Environmental
Protection Aqency for their cooperation and suuoort.
-------
ABSTRACT
The polybrominated biphenvls (PBBs) and polybromiriated biphenyl oxides
(PBBOs) are brominated hydrocarbons that have been used commercially as fire
retardants. PBBs in particular are highly lipotropic and are not readily
metabolized. As a result, they remain in adipose tissue for extended periods.
In Auaust, 197R, a comprehensive medical evaluation was conducted on 42
out of 96 workers frorr. a plant that manufactured decabromobiphenyl and deca-
bromobipheny1 oxide and on 96 control workers from neiuhborhood industry not
involving PBB. The exposed workers were emploved in a plant that manufactured
only brominated products and 96 regular employees worked there in the period
January, 1973 to May, 1978. PRB serum level was significantly hiaher among
the exposed group. An unexpectedly higher prevalence of primary hypothyroidism
was found amonq the PBB workers than among the controls. Their sural sensorv
and peroneal motor velocities were significantly reduced when contrasted with
controls. However, no significant dernatoloqical, neurological or other
adverse health effects were clinically demonstrated.
K
i
-------
HEALTH ASSESSMENT OF OCCUPATIONAL EXPOSURE TO POLYBROMINATED BIPHENYL
(?3B) AND POLYBROMINATED BIPKENYLOXIDE (PBBO)
ANITA K. BAHN, ORA BIALIK, JACQUELINE OLER, LORNE HOUTEN, EMANUEL LANDAU
INTRODUCTION
Brominated aromatic hydrocarbon compounds manufactured for use as flame
retardants have been recognized in the last decade as a potential source of
environmental contamination.^ Polvbrominated biphenvls (PBBs) and congeners
have a close structural relationship to the widely prevalent polychlorinated
biphenyls (PCBs). P3B is a highly lipotropic substance that is not easily
metabolized or biodegraded.^ As a result, PBB remains in human adipose tissue
and in the general biosphere for extended periods of time. Polybrominated
biphenyl oxide (PBBO), an ether, is less inert and slightly more readily
metabolized than PBB. PBBO is still very stable, because the oxygen bonds are
not easily broken.
PBBs have not been associated with cancer in humans. This may be due to
the fact that PBB manufacture first began in 1970.^ Both PBB and PCB have been
associated with the development of neoplastic liver nodules in rats.4 A
possible relationship between PCB and cancer in humans was noted in a small
cohort of workers.
Following the accidental contamination of dairy cattle feed in Michigan
in 1973 by the fire retardant BP-6 Fire Master, principally a hexabromobiphenyl
animals showed anorexia, hair loss, skin roughening, abnormal hoof growth,
liver abnormalities, decline in milk production, loss of weight and death.6
An increase in stillborn and malformed calves also was seen. Animal experi-
ments with PBB and related substances have demonstrated thyroid and liver
changes, infertility, retardation of intrauterine growth and other fetal
effects.The symptoms of possible PBB-related illnesses reported in human
who consumed the contaminated animal product included muscle pains, fatigue,
weaknesses and sleep disorders, skin, neurologic and gastrointestinal
difficulties and immune dysfunction.^-4-16
The effects on human health of exposure to PBB have been studied in
quarantined fctm residents, consumers of farm products, chemical workers
(from Michigan) and in a control qroup of farmers from Wisconsin. Differences
in symptoms were found between exposed and nonexoosed groups; however, symptom
oreva'lence rates had nc consistent relationship to bodv burdens of PBB as mea-
1 7—18
sured in serum or in adipose tissues. No consistent differences in objec-
tively measured variables between exposed and nonexposed individuals
have been resorted. Results of studies on immune dysfunction, for exarrole, hav
1 8-* 19
been conflicting. '
-------
2
PBBs, specifically the octa- and decabronobiohenyls, have beer, found in
the environment (soil, water, sludge) around the sites of certain maior pro-
70 '
ducers. While there is concern with the effects of PBBs on the general
community, workers engaged in their manufacture may have had an earlier expo-
sure to higher concentrations than the community; therefore, health effects can
best be elucidated in workers as the first approach to risk assessment in man.
Workers engaged primarily in the manufacture of PBB would appear to represent
the ideal study croup. Accordingly, a cohort of workers in a plant engaged in
the manufacture of decabromobiphenvl was identified, and a sample of these
workers was examined in August, 1978.
METHOD
The major objectives for the health effects study were as follows:
(1) to measure the extent of worker exposure to PBB;
(2) to determine if associations exist between certain adverse health
effects and exposure to PBB;
(3) to identify early indicators of adverse health effects as associated
with exposure to PBB;
(4) to compare current health status with preexistinq health conditions
and to identify other toxic exposures in order to assist in evaluating
which conditions may be the result of exposure to PBB.
To fulfill the above objectives, the studv was designed to:
(i) maximize comparability of the data collected with data from other
human and animal studies of exposure to PBB;
(ii) examine effects of other toxic substances at work and at home;
(iii) oerrr.it adjustment for age as a known/suspected covariate in adverse
health conditions;
(iv) take into consideration other demographic characteristics, health
conditions and life style prior to, during, and after exposure to
PBB.
THE STUDY POPULATION
a) Expose:! Workers
The approach to the epidemiological investigation oc PBB-induced health
impairments involved both current and past workers. This appeared to maximize
the research potential because subjects are r.ot drawn only fron those currently
employed at the facility. The enlarged samnle included those workers who nav
have been hospitalized or who quit their jobs as a result of intolerance to
exposure in the workplace.
-------
3
The company chosen for the study specialized in the production of
brominated chemicals. Approximately 200 workers had been employed at t.he plant
between January 1, 1973 to March 31, 1977, during which decabromobiphenyl (PBB)
was manufactured. Since then/ decabromobiphenvl oxide (PBBO) has been the
major manufactured product. This plant was selected for study because 1) it
had only manufactured brominated products which minimized the oossibilitv of
potentially confounding industrial chemical exposures, and 2) a hiqh deqree of
cooperation between the investigators and management existed which was an
especially important consideration since the plant is not unionized.
A total of 92 workers (excluding 4 known deceased) were employed at the
plant for at least 6 weeks (or 240 hours) between Januarv 1, 1973 and Mav,
1978. These workers were contacted by registered mail with the assistance of
the company.
b) Control
It was determined that the control groups should consist principally of
blue collar workers employed in the same community thus controllinq for socio-
environmental factors. The study would be conducted in a double-blind fashion.
It was also desirable to draw the comparison groups from other industries in
order to control for the possibility of unforeseen health effects specific to
a given industry. Furthermore, the relatively small size of the PBB worker
cohort suggested the need for a larger control population.
For each person in the PBB worker cohort, it was planned that one control
worker would be matched from each of two selected industrial groups bv sex,
race and ace (+ 5 years). Those used were the 1) local steel workers and 2)
electricians. It should be noted that other groups who were unlikelv to have
been exnosed to any chemical also had been approached. However, factors per-
taining to schedulina and employer cooperation resulted in the inabilitv to use
other industrial Qrouos for control purposes. A supplemental group was
examined on-site and included to complete the age-matched triads.
QUESTIONNAIRE
The interview questionnaires wore constructed to obtain the maximum
information within the time available. Where possible, items from standardized
questionnaires were utilized. Content, format, and language were selected to
reduce ambiguity and elicit maximum response rates. Internal checks on the
reliability of key items included the duplication of some questions in different
portions of the questionnaires. Thus, questions pertaining to exposure to
chemicals, fumes and dust were asked in an oper.-ended format in the lifetime
occupational history as well as in the closed format. The latter check list of
specific toxins was asked after the occupational history had been taken.
Following completion of the field studv, two workers were dec lard ineligible
because they did not meet the criterion for the period of employment. The
cause of death is shown in Table XIV.
-------
4
Validity (true positivity) was enhanced by obtaining information on
symptoms that were moderately severe and cf long duration. Also information
was requested for: (1) dates and names of physicians who were seen for con-
ditions or who prescribed medications and (2) dates of all hospitalizations
including names of hospitals and reasons for hospitalization.
Presented below are the iteir.s contained in each instrument:
a. Demographic-occupatior.al-smoking-drinking questionnaire
1. Aqe, race, sex
2. Educational attainment
3. Places of residence since 1972
4. Lifetime occupational history; specific chemical, dust and
fume exposures, by duration and type of contact (e.g.,
inhalation, skin)
5. Avocational contacts with chemicals, dust and fumes, by
frequency and duration of contact
6. Smoking history
7. History of alcohol and "street" drug usage.
b. Medical history questionnaire
1. Kane/address of personal physician
2. Self-rating of overall health and change in health over time
3. Historv of hospitalizations including admission date, length
of stay, name/address of hospital, presenting problem and
current status of problem
4. History of physician-diagnosed medical conditions by system,
including date of original diagnosis, attending physician,
status of condition, prescribed medication(s)
5. Symptom review (over last five-year period) by system
including dates of occurrence, physician contact and current
status
6. Supplementary dermatological symptom and medical history
7. Supplementary respiratory symptom and medical history
(based on the Medical Research Council Schedule)
8. History of medication(s) usage by drug type, including
specific drug, dates of usage and prescribing physician.
PHYSICAL EXAMINATIONS
a) General Physical Examination
This examination consisted of observations on general appearance, head,
eyes (no.n-f undoscopic) , neck (thyroid), lungs, abdomen, genitalia, lvmnh nodes,
joints. Measurements of blood pressure, pulse, height, and weight were taken.
-------
5
b. Dermatologic Examination
Like other halcgenated aroir.atic hydrocarbons, PBR is stored in the sebaceous
glands of the hair follicles. Exposure to this substance may cause haloacne
dermatitis, palpebral swelling and ocular discharge, and other dermatologic
conditions. To assure that no adverse dermatologic effects would be missed, a
consultant dermatologist was employed to examine all screenees. He recorded
his results on a form modified after Selikoff.2^
c. Neurologic Examination
The reDorts from the Michigan PBB studies indicated the occurrence of
health effects such as fatigue, sleep problems, personality change, lack of
coordination, and other ill-defined central nervous system manifestations
(11,14,15,22-29). Hence a consultant neuroloqist was employed to conduct a
neurologic examination of the cranial r.erves, reflexes including abnormal
reflexes, and to test motor coordination. In addition, the neurologist
administered a standardized "mini-mental status" examination developed by
Folstein and McHugh (University of Oregon).
d. Nerve Conduction Test
A major decision concerned how to measure in the field any possible early
damage to peripheral nerves. To test clinically susnected neurooathv,
conduction studies involvi ng the insertion of special needles constitute a more
reliable and valid procedure. Neurologic experts agreed, however, that under
field conditions and as a screening examination, the use of surface-electrodes
to measure amplitude and velocity of nerve conduction would be adequate. Fur-
thermore, it would have been difficult to justify any invasive procedure without
evidence of neurooathy, and such a procedure might have lowered the participation
rate.
A portable instrument was used to study the following nerves: peroneal
motor, sural sensory, ulnar sensory, median motor and median sensory.
LABORATORY TESTS
a. Markers of Amount of Chemical Exposure - Serum PER, PBBO and PCB Levels
Because of its lipotropism, the best indicator of PBB exposure is the level
of PBB in adipose tissue. However, obtaining a fat specimen requires an inva-
sive procedure (needle aspiration or incision). Such a nrocedure could not be
justified at this time fcr the following reasons: serious health effects to
PBB had not yet been demonstrated in this population; under field conditions,
an operative procedure would be difficult as well as costly; the participation
rate of PBB workers would suffer and the procedure would be unwarranted for
non-PBB workers.
-------
6
A stronc correlation (r = 0.98) was found between fat and serum PBB at
detectable levels in the serum (40% of the Michigan sample).30 Therefore, it
was decided to assay serum for levels of PBB. At the same time, serun levels
of PB30 and PCB would be routinely assayed.
b. Urine Tests of Liver Microsomal Enzyme Induction
PBB and related compounds have been reported to induce a number of
microsonal enzymes in the liver and other organs.^ Induction of two major
enzymes, UDP-glucuronvltransferase and hepatic cytochrome P-450, can be mea-
sured by increases in porphyrins and glucaric acid in the urine. The specific
measures of porphyrin metabolism tested were: total porphyrins (coDro- and
uro-) , porphobilinogen, delta aminolevulinic acid (AALA). Creatinine levels
were recorded as well.
Because of the possibility of diurnal variation, optimal measurement is
made from 24-hour urine specimens. There were several deterrents to routine
24-hour urine collection in this field study: (1) confidentiality of worker
participation in the study prevented urine storage during the working day at
the PBB olant; (2) half of the PBB cohort were former workers; both this group
and the electricians worked at a variety of sites; (3) some screenees arranged
appointments the day before the examination.
Arguments in favor of spot urine collections were: (1) NIOSH reported
(personal communication) that standardized (normal) spot urine porphyrin/
creatinine ratios did not indicate marked diurnal variation; (2) It has been
shown that glucaric acid/creatinine ratios derived from spot urines correlated
with ratios derived from 24-hcur urines,31 3) Group comparisons of PBB workers
and controls would be made.
The compromise decision, therefore, was to collect suot urines from all
subjects during the examination (the amount of urine augmented by ingestion of
r.on-caf feinated suqar-free beverages) and to encourage the collection of 24-
hour urine, either before or after the examination, by payment of an additional
participation fee.
c. Other Tests of Effects on the Liver and Fat Metabolism
Liver damage such as liver enlargement, and fatty swellinq was reported to
be associated with PBB exposure; however, no laboratory tests specific to
effects of PBB as differentiated from excess alcohol or other toxins could be
identified. Therefore, standard laboratory measurements of liver functions
were carried out: LDK, SGOT, SGPT, VGT, bilirubin, alkaline phosphates. In
addition, questions were asked about history of jaundice and the size of liver
was noted at the examination.
P33 is also reported to induce hyperlipidemia. While both serum cholesterol
and triglycerides were measured, the triglyceride value in particular is not valid
since it was not feasible to obtain a 14-hcur fasting blood specimen. (A
two-hour ^reexamination fast was requested, however.)
-------
7
d. Tests of Infertility
Damage to the interstitium (Levdig cells) of the seminiferous tubules is
manifested by impaired maturation of the spermatocytes. As a result, a decrease
in the sperm count will occur before levels of testosterone, LH or FSH are
significantly affected. Collection of sperm samples was not considered feasi-
ble for the following reasons: (1) practical difficulties in collecting sperm
samples on site (personal communication, University of Illinois study of the
workers); (2) sperm specimens could not readily be brought to the examination
site by workers coning directly from work; (3) required abstinence from sexual
intercourse for at least 48 hours prior to examination miqht deter participation
(4) sperm counts are extremely variable, necessitating at least two specimens
to confirm impaired fertility and (5) under field conditions, only sperm count
and sperm anatomy could be determined but not sperm motility from specimens.
Since a below average testosterone level found in a preliminary study
provided some cause for concern,32 serum levels of LH, FSH and testosterone
were measured. Additional information on fertility was collected; inspection
of testes for evidence of atrophy; a fertility history of the workers was
taken; and a detailed reproductive history was planned for their wives at a
later date.
e. Thyroid Function Studies
To measure thyroid function, the following tests were planned for each
screenee; T^, resin T3 uptake, TSH. T3 levels were measured when hyper-
thyroidism was indicated and serum thyroid auto-antibodies assayed when hypo-
thyroidism was found. Also, the thyroid was examined and symptoms of thyroid
dysfunction ascertained.
f. Miscellaneous Laboratory Tests
Routine hematology and an SMA-24 screen were done to detect a wide range
of biochemical abnormalities.
g. Reliability Control
To test reliability, a 10% random sample of all values anc of certain
abnormal values was independently re-assayed. (Frozen serum is stored at the
University of Pennsylvania and at Research Triangle Institute, Inc., in the
event that additional tests are justified.)
h. Procedures
Table 1 in the Appendix summarizes the laboratory procedures.
-------
8
DESIGN OF THE FIELD STUDY
a. Exposed Cohort Grouo
A coded list of all 92 former and current PBB workers employed since
1973 for a minimum period of 6 weeks (or 240 hours), as of May 1978, was
furnished by the company. The company had stipulated that it could not re-
lease the names of employees without individual consent. Therefore, the
furnished list omitted actual names and contained only cohort number, date of
birth, employment starting date, termination date (if not currently employed)
and job classification. To meet confidentiality requirements, the company
sent, bv certified mail, letters to all workers on the cohort list with the
provision that mail receipts, deleting identifying information but substi-
tuting cohort list number, be shown to the University of Pennsylvania.
Letters to twelve cohort members were stated to be stamped "undeliverable".
Cohort members were sent a letter which briefly described the health
effects studv and outlined the types of medical examinations and tests to be
performed noting possible risks. It assured confidentiality in handling data
obtained and the right of the participant to withdraw from the studv at any
time. Workers were informed of the time and location of the study and of the
compensation coverage and amount. Phone numbers of the Principal Investigator
(with reverse charges) as well as of a local nurse (R.N.) were provided.
A detachable Informed Consent Form provided several options for
participation: (1.) medical examination (with knowledge bv the company) and
authorization to release comoanv work and medical records,- (2) medical exami-
nation but without authorization to obtain company held records; and (?,)
agreement for release of company records but without participation in the
medical examination. If the worker did not wish to participate, no replv
was requested. A stamped envelope addressed to the Principal Investigator
was enclosed for returning the signed Informed Consent.
Prior to this mailing, all consent forms were numbered to correspond to
the cohort list number, er.ablinq a later comparison between respondents and
non-respondents. Upon receipt of the consent form by the Principal Investi-
gator, the nurse contacted the worker bv telephone and arranaed for an exami-
nation appointment. She used a standardized telephone format and scheduled
appointments on a random basis.
The week before the examinations, a second mailir.a (r.on-registered nail)
was sent by the company to all members of the cohort to recruit additional
part icipants.
b. Comparisons
The general procedures outl ined above also pertair. to the enlistment of
the controls except for some modifications. For one conr.ro': aroi:p, a membership
roster was provided by a local electrical workers union. The 1C members were
-------
9
sent a letter directly from the Principal Investigator by regular mail. The
detachable consent form did not include the options for the release of comnany
work and medical records. Before scheduling appointments, the nurse queried
workers as to whether they were wiremen or linemen; the latter group was ex-
cluded from participation because of hypothesized exposure to PCBs used in
large capacitors or transformer coolants.
The steel workers were contacted by union representatives. Agreement to
distribute letters and consent forms was obtained at a local union meeting,
attended by the University of Pennsylvania staff.
Additional controls were recruited by personal contact at a county
planning board and similar procedures.
c. Procedures to Maximize Participation
The field study was planned to assure maximum participation. The
scheduling of appointments was flexible; early morning, daytime and evening
hours could accommodate diverse scheduling requirements. As stated, invasive,
uncomfortable and excessively inconvenient laboratory procedures were elimi-
nated from the study (e.g., obtaining sperm, and adipose specimens, etc.).
Inducements for participation included monetary incentives, ($25* plus travel
expenses, with an additional sum of $5 offered for 24-hour urine collection),
the opportunity to learn about their own health and to participate in a sig-
nificant research endeavor. Furthermore, the participants could decide the
extent of their involvement (University access to company-held records,
donation of 24-hour urine specimens, etc.). Finally, the site selected for
conducting the study was convenient, accessible, and neutral.
d. Protection of Subjects' Rights and Right to Information
The principles of informed consent arid confidentiality and the subject's
access to his examination results were strictly followed.
e. Special Procedures for Avoiding Bias in the Study
To minimize possible bias, the studv was conducted without any publicity
and with concurrent controls. The following precautions were taken to reduce
potential bias resulting from knowledge of PB3 exposure or its effects:
a. Statements to ?3B workers and control workers described the investi-
gation only as a study of health effects of various occupations in
the area.
b. Controls were chosen from the same local area so that they could
participate in the same manner as the PBB workers; i.e., concurrent
controls.
^Subsequently, the company announced to current workers that an additional $25
would be provided to participants (all payments were made by the University
to protect confidentiality).
-------
10
c. Appointments were scheduled in a generally random order using a
standardized telephone format; workers were interviewed and examined
in the same type of white gown; workers were informed verbally and
in writinq not to divulge their occupation except in the demographic
interview; medical history interviewers were instructed not to in-
quire about occupation, medical history or symptoms; and all phases
of the interviews and examinations were conducted by different
individuals.
f. Training of Interviewers
The team of interviewers consisted of two medical students, four graduate
students in epidemiology, one nurse (R.N.) and one graduate student in
journalism.
These interviewers received specialized training in the administration of
either the medical history or the demographic questionnaire based upon their
educational background and experience.
They participated in two training sessions. The second training session
was conducted at the actual site the day prior to the actual interviews and
examinations. Interviewers reviewed the final questionnaires and role-played
the respondent, alternating interviewer and respondent roles.
Interviews and examinations were conducted during the last week of August
1978, at the Parish Center. This site was selected because of its proximity
to the PB3 plant and its familiarity to the community. The physical layout
provided ample, well-defined space for conducting the separate examinations,
tests and interviews. Figure 1 shows the first floor of the center, indicating
the location of each examining/testing station.
g. Conduct of the Field Study
Participants first checked into the reception area where they were
greeted, "logged in" on a roster log sheet, and the completion of their con-
sent forms verified. Participants were then given a general description of
the procedure to be followed and handed their master folder. This folder
contained all interview and examination forms (each differently colored and
lettered), consent forms for the nerve conduction examination, and permission
for examining specified records. A control sheet was stapled to the face of
the envelope.
All material in the folder was prenumbered with the ID (serial) number of
the screenee. Four detachable labels with ID number and the screenee's initials
were also provided. The first label was placed on the examinee's gown to
assure correct matching with the forms, the second on the refrigerated urine
bottle, the third on an empty folder given to a Control Station and the fourth
given to a clerk for preparation of labels for laboratory tests. The master
folder was carried by the interviewee throughout the examination-interview
and "emptied" of its contents at the appropriate stations.
-------
FIGURE 1.
Floor Plan of Parish Hall
Kitchen
WAITING AREA
Neurologic
Exam
Medical
Interviews
Demographic Interviews
WAITING AREA
Statistical Check Station
WAITING AREA
WAITING AREA
General
Medical
Exam
Women's
Room
leception
Desk
Men' s
Room
Lab
Refrigerator
Stairs
i Dermato-
loaic Exam
Exit
Interviews
Nerve
Conduction Tests
Screen
Screen
MAIN ENTRANCE - GLASS DOORS
-------
12
After checking in at the reception station, the screenees went to a
dressing room area where he/she changed into an examination qown and gave a
urine specimen. Subsequent assignment to a particular station proceeded on
a random rather than a predesigned, sequential basis. This was necessitated
by the variable length of time required to administer the demographic and
medical history questionnaires. For each phase of the study, the initials of
the staff interviewer/examiner responsible and time 'in and out) were noted
on the Control Sheet affixed to the screeneo's folder. In this way, phases
yet to be completed by the screenee could be readily determined. All com-
pleted forms were taken by the staff to the statistical check station where
they were reviewed for completion and assembled in the code-numbered empty
folders.
Prior to leaving the examination area, screenees were asked to give their
final urine specimen and to participate in an exit interview which consisted
of a second check that all phases had been completed; "debriefing"; the signing
of authorizations for obtaining medical records and consent to give a 24-hour
urine (if not already done), and the payment of participation fee and travel
expenses.
h. Post-field Debriefing
On September 12, 1978, project staff met with the field interviewers to
evaluate: a) quality of interviewing process; b) congruence between written
protocols and actual process in order to adjust for any noted deviations; and
c) extent to which any alteration in procedures might, have produced biases in
information collected.
Based on the reports of the interviewers, it is believed that there was a
hich degree of congruence between their activities and planned procedures. It
appeared that problems encountered in the field {e.c., scheduling) did not
influence in any significant direction or degree the interviewing process it-
self. Additionally, a high degree of rapport between the interviewers was
reported by the latter. These fin dings, corroborated by a nationally recognized
consultant's on-site report and corrjr.ents of individual screenees to the Prin-
cipal Investigator, enable project staff to have confidence in the "face"
validity of the information obtained during the interview phase.
i. Data Processing
As in the field study, the same concern for "blinding" was followed in all
data processing phases. Discrete duties were assigned to different individuals
to minimize any opportunity for decoding employment status. Thus one individual
worked only with workers' permissions and copies of company-held work and health
records. The questionnaire/instruments and laboratory reports were coded with-
out reference to emoloyment status, with the exception of the demographic occu-
pational questionnaire which provides employment information for coding. Prior
to any statistical analysis, data were cleaned from punching and transcription
errors. All laboratory values were assayed in ceded form and their transcription
dcublv checked.
-------
13
j. Publicity
Since this study had almost been aborted by the quoted statement of the
ther. EPA Project Officer on January 17, 1978 to the local newsnaoer that "sug-
gestive evidence" exists that PBBs are carcinogenic, special efforts were made
to insure that all further Dublicity was minimized and all press relations
were handled through the University.
METHOD OF ANALYSIS
The analysis began with a general description of each of the tested
variables for the total study population. This preliminary phase was important
for the determination cf the overall distribution of the variables and identi-
fication of outliers. It was then possible to determine the reauired trans-
formation necessary to permit application of linear models techniques or to
demonstrate the need for the application of ncn-parametric tests cf significance.
Frorr. examination of the distributions, a number of laboratory tests were found
to be skewed to the right; a logarithmic transformation was nade to improve
normality.
Appendix Table TI lists the "normal" ranges of all laboratory test values
given by the laboratory or the literature. This table also gives the mean,
the standard deviation and corresponding range for 95 percent of the study
population means. It is evident that in a few laboratory tests, (e.q.,
leucocytes), the normal ranees as reported in the medical literature vary from
the 95 percent range found in our study. This deviation can be explained by
either some variation ir. the laboratory techniciues or the effect of handling
the specimens during the phases of the studv. However, it is expected that
the samples were handled in the same manner in both the PBB exposed group and
the comparison controls. As a result, any statistical comparison described
is believed to be valid.
One hundred and thirty-eight workers particicated in the study: 42 PBB
workers, 42 electricians, 26 steel workers and 28 others. Tt was nlanned that
all PBB workers would be matched by age group, race and sex with the workers
from the comparison groups. However, under field conditions, it was not
nossible to "control" for aqe (e.g., not all oersons who made aooointmer.ts
came in for an examination). The electricians, in general, were older and the
steel workers vounger than the subject group. Therefore, since many of the
biological variables studies are aqe-related, age marched triads were formed
for males (Appendix Table III). In forming the triads, 2P of the PBB workers
were matched (+ 5 years) with ~he electricians and the R voungest PBB workers
were matched with workers from the "other" qrouo. All 25 steel workers were
matched with the PBB workers; the renaininq 11 PBB workers were matched with
workers from the "other groups". 3xcent for matching by nearest ace and
occupational group, selection was made without examination of other
character:sties o^ the individual.
In the preliminary analysis, the study population was grouped in five ways:
1. the total population (138) of examinees
2. by triads - 36 male PBB workers matched by aqe and race to controls
(electricians and others, steel workers and others) (see Table IV)
-------
14
3. by sex and employment
i. 39 PB3 workers (36 males, 3 females)
ii. 26 steel workers (25 males, 1 female)
iii. 42 electricians (42 males, no females)
iv. 28 others (24 males, 4 females)
3'. as (3) with medical exclusions (alcoholism, diabetes, etc.)
4. by PBB male workers versus all other male controls (36 PBR workers
vs. 91 controls)
4'. as (4) with exclusions as in 3'
5. by triads, sex and race (10 groups)
i.
A
34
- PBB workers, white males only
ii.
B
34
- electricians and others, age-natched with A
iii.
C
34
- steel workers and others, age-matched with A
iv.
F1
3
- PBB workers, females
v.
F2
5
- control females
vi.
KM1
2
- PBB workers, non-white males (1 black, 1 other
minority)
vii.
MM 2
2
- non-PBB workers, non-white males (1 black, 1
American Indian)
viii.
EXl
16
- unmatched electricians
ix.
EX2
5
- unmatched others
X.
EX 3
3
- ex-PBB workers who were classified as ineligible.
3 ¦> .
One-way analysis of variance tests (ANOVA) usinq BMDP7D " identified
significant differences with respect to laboratory values between the exposed
groups and the comparison groups (unadjusted for the effects of other
exposures). The following points are noted:
a. A one-way ANOVA among the triad matched groups can be considered as
an analysis of covariar.ce, where age was controlled as a covariate, without ac-
tually estimating the ane effect. The BKDP7D proqran also provided a detailed
description of each of the biological variates, i.e., histograms, means,
standard deviations and ranges within each group as well as for the total
study population.
b. ANOVA among the four groups as defined in (3) above was apcrooriate
since it was exoected that the "other" grouo might "contaminate" either the
electricians or the steel workers, and as a result, existing differences
would not be identified. 3y applying such methods, it was assumed that aqe
had a negligible effect on the variates studied.
c. Applying the BMDP7D program to the 10 subgroups, as defined in (5)
above, was necessary for completeness.
-------
15
d. The editincr procedure pointed out variables with non-normal
distributions. In such cases the comparison among grouns was carried out by
applying the logarithmic transformation prior to the AMOVA. These variables
are indicated with in Appendix Table II. No transformation normalized
truncated distributions such as the TSH assays. In those cases, non-parametric
methods (i.e., X^ test) were substituted for the ANCVA test.
e. Special analyses were cerformed for the nerve conduction studies. It
was not feasible to maintain age matched traids follovinc the exclusion of all
subjects with known extraneous causes for neuropathy (which resulted in the
groupings 3' and 4' above). The analysis of covariance was an appropriate post
hoc procedure by which to adjust "or possible bias introdaccd by the uneciual
age distributions which resulted. Assumptions which must be met in order that
the procedure by valid were examined, utilizing ?-P plots34 and the exploratory
data analysis techniques due to Tukey."^ The choice of an appropriate trans-
formation must be determined with primary consideration given to the form of the
functional associations of each nerve conduction response with various covariates,
e.g., age, heavy metals and solvent exposures, the PBE blood serum levels and,
for the workers, measures of occupational exposure to PEE such as length of
employment at the plant. Techniques which examined, in order to improve, the
applicability of the classical multiple linear model for the analvses of these
variables were implemented subsequent to the preliminary analysis reported here.
Further consideration of appropriate transformations for the peroneal motor,
median motor and median sensory velocities remained to be made and are to appear
in a separate paper.
The above methods apply to quantitative variables. Qualitative variables
were analyzed by chi-square tests. In the detailed analysis of thvroid
function data, the Z test, Fisher's Exact Test and the Mann Whitnev Test were
employed.
In summary, investigation began with the description of the sampling
distributions of measured variables for the studv as a whole. For the Quanti-
tative laboratory variates, analysis of variance F statistics were examined to
determine if there were significant differences across the followinq selections
for grouns: (2) the triads, (3) by employment (PBB, steel workers, electricians
and others), and (4) by PBB workers and all controls taken together, "he major
comparison presented here will be for the matched triads, (2). Further pre-
liminary study of the nerve conduction velocities included analyses across
groups defined by emnlcyment (3*) with adjustment for age using an analysis of
covariance model.
For the comparisons across triads, ANOVA F statistics, testing the overall
adequacy of the model to explain the variation in the responses, are reported
with the P-values appropriate to two-sided tests. An a = .05 level of significance
would be consistent with standard practice in assessing the reported P-values.
For the laboratory variates, no attempt at contrasting specific controls with the
PBB group was made, and thus no determination of simultaneous testina levels
(with an overall a level of protection) was necessary.
-------
16
However, for nerve conduction studies, reducer1 velocities were the only
alternatives possible. Thus, one-sided tests for contrasts of PBB workers with
controls were indicated and P-values (indicated bv ?*) appropriate to one-
sided tests are qiven. No overall, or simultaneous a-levels are attempted for
these contrasts in this exoloratorv investigation.
It should be noted that in addition to the contracted computer analysis,
a ohvsician reviewed selected ocrtions of the Questionnaires, e.a. , svinntoms,
and the results of the physical, dermatolocical and other examinations.
RESULTS
a. Demograohic and Occupational Characteristics
One hundred thirtv-eight workers p
42 PBB workers: 34 white ir.ales, 2
(3 white males were ineligible for
1 worker who participated was not
42 electricians: 42 males
rticioated in the field stndv:
non-white males and 3 females
the study based on employment data and
on the original cohort list)
26 steel workers: 2 3 white males, 2 non-white males and 1 female
28 others: 24 males and 4 females.
The 39 PBB eligible workers represent approximately 45 percent of both current
and former workers who were emnloyed by the company for at least six weeks
during the studv period (January 1973 - August 1979). The studv included 3
female workers enoloved by the conoany from 1972 - 1975, 1965 - 1973, and
1975 - 1978. Because of their small number, the characteristics of the women
are described separat.elv. Fiqure 2 documents period of employment for the 36
PBB male workers.
The median age of the 36 PBB males was 30 years (ranoe 22 to 57 vears).
PBB workers did not differ sicnificantlv in age cron the entire Tale cohort.
The mean duration of total employment (27 months) and of em.plovnent during
PBB production for the PBB workers (19 months) (Figure 2) did not differ sig-
nificantly from that of the entire male cohort. (z value fcr total duration of
employment: -.644; z value for em.plovment during PBB production: .058.)
b. Medical History
The frerruencv of reported symptoms "anv time during the last five years"
by svstem and occupational group (Appendix Table V) indicated no statistically
signficant difference in comparing the entire three major occupational groups.
/
-------
- +
-I 1-
-I 1 {-
4 k
-I 1 1-
1/79
1/78
1/77
1/76
1/75
1/74
1/73
1/72 "¦
1/71 --
1/70 -
1/69
-i 1-
+ --
Figure 3. Period of Employment of 36 PRE
Workers
Length of Employment
+ *
Date Date
work work
began ended or
+ August 1978
+—*-
-I 1 1 V-
-i 1 I-
-fPeriod of PBB Production] ~
ro
+ --
-I 1 1 1 1 1 1
-I 1 1-
1/69 1/70 1/71 1/72 1/73 1/74 1/75 1/76 1/77 1/78 1/79
Date Work Ended or August 1978
-------
c. Physical Examination
Two PBB workers and one electrician had an enlarged liver (defined as
>13 cm. total or 2 cm. below the right costal margin). Review of the medical
histories of these nen revealed no indication of antecedent liver disease.
All three had at least one elevated liver enzyme test. Of these, only one
person (PBB worker) had a history of heaw drinking.
d. Dermatologic Examination
The data were reviewed for any evidence of haloacne dermatitis. Active
acne, or scarring resulting from a prior acne condition was evidenced in 69.2
percent of the PBB workers, 65.4 percent of steel workers, and 76.2 percent of
electricians. The difference is not statistically significant.
e. Neurologic Examination
Neurological problems were clinically identified in only a few workers in
the studv. Comparison of the four groups (PBB workers, electricians, steel
workers and others) showed no statisticallv significant difference (see Table
VI). No abnormal motor and coordination behavior or abnormal mental state was
demonstrated by the "Mini-Mental Status" examination.
f. Laboratory Tests
The mean values of the quantitative variables studied are presented in
Appendix Table VII. The percentage distributions of laboratory values, defined
as above-normal, normal, and below-r.ormal as found in the matched triads are
presented in Aopendix Table VIII. Abnormal results were defined by the labora-
tory or the literature when such data were not available, e.g., spot urine
values. A "normal range" was determined for the study based on the marginal
distribution of the study population. Improper collection or storage of some
specimens may have resulted in some of the abnormal values.
1. Hematology
No statistically significant differences were found in the three age-
matched groups using analysis of variance techniques (Appendix Table VII).
The laboratory normal range differed from the 95 percent value for a number
of these variables (e.g., LDH) (Aooendix Table II).
2. Blood Chemistry
A statistically significant difference in the mean for the three groups was
shown for serum phosphorous (P=0.03) (Table VII). Statistically significant
(P=0.03) negative correlation between length o^ employment at the PBB factorv
and particular serum levels (e.g., total protein and serum trialvceride) was
found (ADDendix Table X). The findings of high serum phosphorus in electrical
workers and of a negative correlation between raeriod of emplovment and total
protein and triglycerides may be interpreted as chance associations until
additional suooortive information is provided.
-------
19
3. Urine
Appendix Table IX indicates that the qualitative tests on soot urine
(nitrate, pH, orotein, glucose, ketones and blood) did not differ signficantlv
by occupational group.
Liver function tests results of all examinees were studied in detail.
Outliers above the values for the study population on the following liver func-
tion tets were studied: YGT, SGPT, SGOT, bilirubin, LDH, and alkaline ohosoha-
tase (see computer printout of histroarair.s and related data in Figure 3) . Five
PBB workers had elevated levels but only one subject had an elevation on more
than one test. Four electricians had elevated tests including two Dersons with
more than one elevated test (Table XI).
Table XI lists for the individuals with elevated liver function tests other
laboratory findings, medical histories, physical examination, and habits in order
to explore a possible etiology of the abnormal tests. Of the five PBB workers,
two admitted to moderate to heavy use of alcohol presentlv or in the cast. Of
the four steel workers, one was a heavy drinker, and two had serum hepatitis.
Of the four electrical workers, one admitted moderate to heavy drinking and
another was being treated for lvmphorr.a.
Due to the small numbers and other possible explanations for the liver
function tests, an occupational exposure accounting for these abnormalities
cannot be postulated.
4. Endocrine
a. Thyroid function
The study documented four cases of hypothyroidism in 35 men* employed in
the production of PBB/PBBO but none in the 69 control subjects.^® The findincs
of an elevated TSH and low free thyroxine index indicate Drimary hypothyroidism;
that is, disease of the thyroid gland rather than of the oituitarv or
hypothalmus-
In the documented cases, the four men who had an elevated serum TSH had a
low or borderline low serum. T4 and free thyroxine index. None of the 89 con-
trol subjects (including the 72 control (triad) subjects) had an elevated
serum TSH. Since all subjects were questioned about occupational exposure to
74 toxic substances, it was determined that the three PBB and PBBO workers who
had primary hypothyroidism had only three common exposures: PBBs, PBBOs, and
bromine. The fourth worker who had hypothyroidism was known to have been
exoosed to PBBOs and bromine.
The PBB group had significantly more subjects with elevated serum, concen-
trations of thyrotropin (P=0.00o). The free thyroxine indexes (P=0.06) and
serum T4 concentrations (P=0.11) did not differ significantly. Although more
PBB workers than controls had antimicrosomal antibody titers greater than 1:100,
the difference between groups was not significant
-------
FIGURE 3. Histograms of selected liver function tests for matched triads
yGT
ygt
PBB
(N=36)
Electricians, Others
(N=36)
,.,,4
*
~
* * *
*»*~~»***»«*
Steel Workers, Others
(N=36)
r--1 ni'ii I rs
•isft.O.i.s i
4^0.,1^,1 J
300.non)
JHil\ r\ 1
1 .ono)
VJ'I.OMI
? 70..inn)
? jn ,ooo i
?1 o .o^o >
1 ¦ii.ion i
\ r>,nnn i
i2n.'T>ni*
on.ono» * **
f,'l 'Vl ^
o.oooj*
-*0.1001
_*o.ooo)
-00.non )
GWr»«"P '.HAMS fti*K n ; • n |>" o AY »*s IF 1 HFV OPJ i"C I'^F '• T 1 V f
*
~
» t
v I I » t t * 11 »»**»•****«*
tt
1HP
>' t ft M
S. fir'/.
MA XI
i 1 •! I
17.7??
77 . i?h
*h
in.10!)
j 1 . *M
», £,. o 'i
lo.'T.ri
3P? ,»10C
R .000
J' .l ¦*«
S ¦> . 7 y rt
3^ .ooo
j to .ooo
7.000
fci.i. c-:> "«•»*-*« r >"'j i r o fTAS'-'.s b "¦ s:-'rr j nj., rrrFs r p f i t»,m v a pi am 1i
F 'i
S. nK\
*¦
• '• i I "!
"I : T • 1
3^.7 H
5 1 . » b 7
1 o>J .
»•;>.<
511" OK M I" ! f-\S IF
ir FT. FFN t J») ? . u •' 7-' /
IT,, ir 77* 1si 1.^3 fin 11 5
*1 r»' it t .7.H1O j 07
f /*-i r,r>iiA!-*F F P .1T I n
0.75 97
7J17
'PPO. F r *Cf'F r-f-'i
o.7T10
O
-------
FIGURE 3. Histograms of selected liver function tests for matched triads (cont.)
SGPT
SGPT
"rnn.'i'"! .s
;;o . ion)
7 . o o n )
77. 0 - <»i
ftp .inn i
sn .inn )
S •> .1)10 )
I . l n n i
II .O.lf )
in. iii.i)
)
P.0O0)
7 > >-•)
1 r> . 100 1
17.00-11
in)
i ..-no)
GR U'P -IT A
uk a j
S. '>1 V.
s
•! A < T .111 v
,/ T J J • i' | •/
PBB
(N=36)
Electricians, others
(N=36)
Steel Workers, others
(N=36)
* ***
¦J S IRk '-I-' i'HFP H V
?o.? / a
a . •> 1 I
1 ri . 0 ,1
^ 7 . o o o
1 ? .0"!
~
~
*
«
t *
' * ~
ntMnn»
~»***~*4»~
» ~ * ~ »
IK TKh y C°1 >• C 1 PK ' 17H *'$¦, I 'F- OTH W '. 1 Sk
?"? .«0h 71 .f f>7
i ?. i ? r,. 71 f<
U; . 0 , - (I 3.'.>.0 0,1
7() .orn 7 3. 100
1 7 .oon 1 ?. r>oo
;!.l. J'M'loS rriv^T 'K'1 (TV'K.S F X f" 1,' Mi R D Jf C ! A1, f'l'KS mc t j TMpt VMMM 1,F )
SIH' r'K S». IJ A T- k IK vtAf' S.l'UCK F t'AT]n
0 . f< 7 7 0
nf'/.
'J
•II ¦ I«
• f "I'll.-
?i .°n
Il.i7!
1 ,'1H .
M. l
1 7.0U1
^ r, 1 F'
' I 7 1 '•
rrr r i
771 .
I XL /.n .
7
1 fS
1 r.7
1 7 o .') 4 *
p'-iop. «• rxo^rr.
'i.tint
-------
FIGURE 3. Histograms of selected liver function tests for matched triads (cont )
SGOT
SGOT
PBB Electricians, others Steel workers, Others
(N=36) 4 # _(N=36) + . . . . lN=3£l
v I OP') r'IT 1)
117.00O)
los.o'vn *
ja .n.tn)
'' 1 . 0 o 0 1 *
J. n o 0) «
7 7.00 0) * *
7 •"* . 0 0 0 )
»
¦>'i .''0 0)
4Q.OOOJ <
1 •> .ooo) ~*» »
. o on )*«»~*
?1..10-T ~**»*~
1 1 ,nftri)tt*n* m utt • «»*~»~~*»*~»*»
7 . 0 00 ) ~~ » ~
0 . o o n )
-7 .000 )
coup -'c-as-jj OF-.nr^n >$/ * *s if 1 wvv c r-1 > cii'F • itm ~ f '? r.'i i f^ l sf
4
4
4
* » *
• » « » «
» * * * *
^TA'J
s. 0- '/.
•1
«. eyj-11-
•¦"l • i • • I i • <
) > . n i /i
1 1 . 4 s 7
i 3'>
12
ir- .a0 0
1 7 .0 o 0
1 1 .0-'>r'
A r I r:->nMr>s y > cp, (CArfS X f |..if> I: If-' r-rt-.C I AI corf r> «• H 1 I'f I- VA»- I F 1
'Tv:
S. Mt-V.
• A *T M|v
" y »• y mi-«
1 7-.hMS
n«,
1 "7 ">0
Q . OO0
S11'
rf '( • V> f.
!- T i • 1 r
i • r£ i
r f .s (I • fl f-fF
J S ° . 7 ° r 3
UM3.1»'7()
< ^ r>' ~ . o • in
I F
?
1 0 fS
11.'
r;M «=" p.viir
17y.B:i*| O.5h09
3 1 s . 11 '-J 2
i-M-nM. !•• rxrr-Fnr'
o ,t--:.73
-------
FIGURE 3.. Histograms, of selected liver function tests for matched triads
" ~ ~ T0TAI, BILIRUBIN ~ ~ " ' ' -
TOTAL
ILIRUBIN
PBB
(N=36)
Electricians, Others
(N=36)
I'M':
7
7
7
n>
/
?
7
1
t
1
I
1 .
1 .
I .
f>.
•>.
-0.
r.p<"'Up
i J' rs
. •' f'n >
,7it ! J '
. ri S ">)
, I m 1
» 7 s o 1
. I "0 *
."sn)
J
.sorn
. )
V00 1
i s n)
i o } ~ ~ *
7 50 1 *
^ n (V) * * »
1 Sf> i •-» •
l'n i ~ » »
{ r> o )» ~ ~
n n o l
¦••r av s
~ ~ * *» ~
» ~
'> F '<
' ft A T
f I
.: 1M
.1 /
A Mr, r-.t".
r>. a a h
o.i! 9
1ft.nrn,
7 1/'
n . i n o
01 pn -.- » t * t 4. « »
« y y
rpi '.'pirF ¦ J T!'
0 . <3^
p, .> o o
?f> .poo
0 . ° 0 0
il .0 }<*•
Steel Workers, Others
" (N=36)" ,V"''
*•'5 Cll'KP
*
«
* *
» » *
* *
* * *
* s .~
; « 4
1 St
T * * * * *
tttVt.t ~
* ~ * *
'• ; Vi J
r .'?^4
oon
'•!.! t>l"
M.I. Hips CO'-'.ni t.-'i f£.\sp<> t> sri-ciAL (J f'fit'.5 FOR fitf-pf. VAP-TAM F-)
t n i
¦»KV.
s
XT" l,M
! ' T ' "••
3 MS
V'7
1 f..(.
7.blO
'."in
si'-' .'I*-' St. I'A IF 5 I F
i t • rr" o. ' i s it'
'I t l. 1 f . 1 l a 4 1(7
rr p.-
:i.»JK F R fi TIU
0 . ? S * s
' . f> 7 ~< 2
i;';ii«, k Fxr^K-nf.
^. 1 ? n
-------
tIGURE 3. Histograms ol selected liver function tests for matched triads LDH (COnt)
PBB Electrician*, Others steel Workers, others
' LDH (N=36) (N=36) (N=36)
v I nny IV r.S
7?o.ooo)
000 ) 4
f>n .OOP )
6 1 i.ono •) ~
6oo. ono )
S 7 0 .1) ) *
Ml.'lOi) V
sm.nin)
i>;o. )oo i %
4 s o . o o o )
I7 )»*~~*~** > .•**~*~»*»* vtHtt'MMttt
j.jo.nnoi (**t« MM* ~ *~<.**
710.000)** «»<* »**~
14 0. fino ) * * * *
mi.Tiii t
c; ' 1 1 .'S AN-? !'K 'OTKO Y ~* 'S TF 1 KFY Cr') f. CID f V JTH * 'K , f.'S f"*'HEP'< ISE
M F a u
5. nKV.
V /\ Y 7 < | i .1
I *. r
? » s . 7 s o
« 7 . 7 $
in .ooo
r,?s .oon
1 «7 .000
7^3. 772
*>7.736
3m. 000
SS3. 'inn
1 « 3 . 0 0 0
7«7.139
,0
" ("n.viJr pn.ir i n ^r; (r»,-,t;,c KXf MiO*"'* IF M'FCIM- Cff? K i. I i 7 n j. f- f- VA^l API-F )
SI'.' 0K SOI'M-F.S
IF
- K A '
SVnA"wt F " AT in
F-'.n^. F FX(»TFi
'•FA'1
s. il:;.
'4
y t "im
• 1 1 *-* i 1
7 31.7'M
7 'J . 1 / R
10k'.
^ns .onr)
1 8 7 . 0 0 0
KFT.FF" ?'.7.7ni 7
. J T >' T r r. 7 '< ti .'1 3 . 2 0(V 10»J M 3'.i,cjPn
IT'Tftr, 67M-.90.SVi0 107
lo2
0 '1 0
p
-C
-------
FIGURE 3. Histograms of selected liver function tests for matched triads (COnt)
ALKALINE PHOSPHATASE
ALKALINE
PHOSPHATASE
PBB
(N=36)
Electricians, Others
(N=36)
Steel Workers, Others
(N=36)
•U'n'ni i rs •
1 31.000)
tlo.ono)
11?.ooo )
1 us.oo n
. (>oo 1
"> 1 .iino l
P4 .0'^) l
77.000 )
7 0 .0 o o )
lii.nnj
S . o o o i
4')..) oo i
1 ?. o o o ) ~
* s. i oo)»
. .' 0 0 ) * * t
, o n o )
,no.>) *~ »
.000)
, 0 o-i i
?1 ,
I 1,
7,
.) ,
-'KV^S AHf- RV
~ H U 4
• ** 11« * * }
*4***»*~~~*~**~»
« ~ ~ *
I F T!!"-;y C" I "C I PF ;• Till ~
~ ~
•.<*» + ****«*«********
* ~ *
fTMKP;. ] SF
a 'i
s. D I-' V.
J
"./I X I 'll /
''.I' T |M
2 1 . 0 H 1
S . I-1 '¦>
3 6.0-10
1 o . 0 ; .i n
1S.ooo
?7.)u 4
1 S . 3 h t.
3 n . ("• 0 0
1 .1 q . n r, 0
1 3 .mn
? . I ° 4
4 . 9 1 S
3^.0 0 0
3 . 0 0 0
1 1 . 0 0 0
if t,
f! i- r • K[\ (rs SrS I-" V C* I, I' !' r'O ] F sr^r] "I, CIT« f-
V Ml:
1 i |.r. (. VAM t >( v 1
;¦ F \ >:
s. r>K\i.
•J
*; y j *, ¦-
" 1 ' l >''if
2 S . 1 S 7
9 . H 7
10ft.
"i'I.i!
1 3 . '.'00
r-1>iif M i;rrFS [ F
V--r 'IF t -J ?^'J.VQh3 l
¦ IT, i.: 1 r; 3e .ov-ii 1
,
-------
26
One year after this study was performed, reevaluation of three of the four
workers with hypothyroidism indicated 1) all three still had markedly elevated
thyroid antimicrosomal antibody titers, 2) two still had low free thyroxine
indexes and high thyrotropin values and 3) one had a normal free thyroxine
index and a high-normal thyrotropin. None of these three workers had been
treated with thyroid hormone.
PBBs, PBBOs and bromine must be considered as possible causes of the
hypothyroidism in these workers because the plant produced only PBB and PBBOs
and because PBBs or PBBOs and bromine were the only substances to which all
four hypothyroid men were known to have been exposed. Animal studies describe
a statistically significant decrease in serum T4 concentration in rats after
PBB administration.^7 This response was dependent on both time and dose.
Serum concentration of thyrotropin in these animals was elevated. It is
suggested that a defect in the iodination of tyrosine is a mechanism by which
PBBs produce hypothyroidism in rats. Although studies of thyroid function
have not been performed with PBBOs, thyroid hyperplasia has also been reported
in rats given decabromobiphenyl oxide.
The subnormal serum concentration of T4 five months after cessation of
exposure in rats suggests a persistent effect of PBBs which could explain the
hypothyroidism in two workers who had not been exposed to PBBs for five years
before the study.
To our knowledge, there have been no other reports attributing human
hypothyroidism to PBBs or PBBOs or elemental bromine; however, the development
of thyroid abnormalities in animals given PBBs and PBBOs supports this hypothesis.
b. Reproductive function
Although a few subjects in the PBB study group had elevated LH's or FSH's,
the differences among the three groups were not statistically significant.
Testosterone levels of all PBB workers were within the normal range and no
significant difference in mean values by worker group was found (Appendix
Table VII). Futher investigation of these hormones is needed in relation to
exposure.
4. Urine
Spot and 24-hour urines were analyzed for creatinine, porphyrins,
glucaric acid, delta-aminolevulinic acid, and porphobilinogen. The spot urine
values were analyzed as ratios to creatinine (see Table I). No significant
differences were found in comparing the age-matched triads.
-------
27
g. Nerve Conduction Tests
The following preliminary results serve the general purposes outlined
above, namely for data cleanup and outlier detection, establishing the study's
empirical distributions with means, s.d., ranges, etc., as well as allowing
initial comparisons across a selection of croups.
A first comparison of all examinees (< 108) or. whom nerve conduction
measurements were made showed the following orderings for the means of the
three grouos (triads) comprising workers natched on race, sex and age, where
P: PBB workers, S: Steel workers, and E: Electrical wiremen.
Triads (42, Page 13)
Nerve Velocity MMV PMV MSV* USV* SSV LSSV
Order d Means E4 drinks/day) and 32 with
excess alcohol intake or.lv. The following tables present analyses on groups
comprised of 23 PBB (P), 31 electrical (E), 12 steel (S) and 16 other (O) white
male workers from the area. Of these there are three missing median motor, two
missing peroneal motor, and one missing sural sensory values due to factors at
the time of the examination, e.g., patients requesting that the testing be
stopped. Numbers in brackets refer to group sizes in the analysis following
these exclusions.
The next table reports one-way ANOVAs excluding all subjects with known
extraneous cause for neuropathy. Results of one-way ANOVAs across the four
groups and across the three groups define-" by industrial occupations were:
Four Groups (#3', Page 14}
Nerve Velocity MMV[72] PMV[72] SSV[72] LSSV[72]
Order Means E
-------
28
Three Groups (#3', excluding "others")
Nerve Velocity
Ordered Means
P-Values
MMV [62 ]
E
-------
29
Three Groups Comparison on SSV, PMV and MM"'/
ANOVA Nerve Velocity 5SV[62] LPSV[62]
P-Values .1038 .0858
Ordered Means (Unadjusted) P
-------
30
The extraction and cleanup procedure was validated using aas chromatograDhy/
electron capture gc/ecd as recommended by the Standard Procedures. Serum sam-
oles were spiked at 58 ng/ml with hexabromobiphenyl. Two extracts gave 76 and
75 percent recovery for the entire nopulaticn.
In order to obtain positive identification as well as quantitation, the
technique of gas chromatography/maspectronetry/computer (GC/MS/COMP) was used
for the analysis of the serum extracts. This technique has been used to iden-
tify and quantitate PCBs, PBRs as well as PBBOs in a variety of samoles including
the identification of P3B0 in human hair. Limits of detection for the reduced
sample size ranged from below 1 ng/ml to 25 ng/ml. depending uoon the molec-
ular weight and gas chromatographic retention times. The details of the GC/KS/
COMP and all the laboratory results (using new identification numbers in random
order) are given in the above RTI report.
The distribution of the total serum levels of PBBs, PBBOs, and PCBs, among
the occupational groups (?BB workers, electricians, steel workers, and others)
is described in Appendix Table XII. The number of cases with detectable levels
of each of the chemicals analyzed, as well as the minimum, and maximum levels,
are shown in Appendix Table XIII.
As indicated in Appendix Table XII, a significantly higher number of PBB
workers had a detectable level of PBBs as compared with other workers in the
study (35.9 percent as compared to 12.1 percent, chi-square = 8.67, p<0.005).
Among workers with detectable PBB levels, the PB3 workers had significantly
higher serum levels (p<0.001) as compared to other workers (Z = 3.61, Mann-
Whitney U test). No significant difference in the frequency of detectable
levels of ?3B0s and PCBs nor in the absolute values of these chemicals was
found among the occupational groups.
It should be noted that although the factory concentrated on manufacturing
decabromobiphenyl and decabromobiphenyl ether, there was no positive identifi-
cation of C^2Brig or C^2Bri3°- This finding should be explained by the facts
that not all the studied chemicals have similar persistence and that the hexa-
bromobiohenyl may be more persistent than the decabromobiDhenyl. Also, the
decabromobiphenyl ether would be more easily metabolized and excreted. As
Dr. Renate D. Kimbrough indicated in a letter dated May 30, 1978, "Analysis
for decabromobiphenyl ether may be difficult. The decabromobinhenvl is
probably not absorbed as efficiently by mammals as hexabromobiphenvl."
Apparently only C-^HgBrO and C^^B^O were detected, and in relatively low
dosages.
These findings imply that anv study of health effects related to PBBO
exposure should be based on employment characteristics (such as neriod of
employment, job description, etc.) rather than on ?BBC serum levels alone.
Accordingly, adverse health effects related to industrial CBB exposure should
be discussed in view of both employment data and serum levels.
-------
31
h. Female Subjects
Eight females, aged 21 to 47 years, were studied including 3 PBB fenales
and 5 control subjects. No separate sex-specific analyses were performed due
to the small number of subjects. However, the medical, reproductive and social
histories, general physical examinations, and laboratory results were reviewed
ir. detail. Only one woman, a steel worker, had a significant medical condition,
chronic hepatitis with elevated liver function tests (Table XI). Two women had
had hysterectomies and one was menopausal, explaining their elevated levels for
LH and/or FSK. Two women had had several miscarriages; neither one was a member
of the PBB group.
LIMITATIONS AND SUGGESTED FURTHER ANALYSIS
1. Comparative Analysis
This was an initial, exploratory search for possible adverse health effects
in human subjects occupationally exposed to PBPs. Only univariate distributions
have been examined and preliminary analyses necessarily were concerned with com-
parisons across study and control groups on univariate responses. Since the 36
cases in each triad were age-matched, all analysis of variance tests across the
three study triads that have been reported here (with the exception of the nerve
conduction studies as noted) have controlled for age. Thus, further studies can
be undertaken within a quasi-experimental design in which age-matched group
sample sizes could be projected to confirm, at a preassianed oower, the existence
of suggested effects estimated from this study.
However, it is inherent in the observational character of an occupational
health effect study that biases due tc extraneous, possiblv confounding, factors
(i.e., those not "randomized out") be controlled for via the appropriate post
hoc analysis. For a large number of the responses in this investigation, a
larger than expected, or "normal", range was observed. The alternative hypo-
thesis which postulated that this variation was due simply to occupational ex-
posure to P33s could not be established for almost all of these responses.
Comparisons across study and control populations would be more discerning if
control of extraneous variation due to factors associated with the response whose
distributions differ across treatment groups, were attemoted.
Further, various multivariate procedures could be pursued for the numerous
laboratory responses which exhibited a wider than expected scread. An initial
inquiry into possible correlated factors has suggested variables which should
be investigated via multivariate methodologies, e.g., lead exposures correlated
significantly with several nerve conduction velocities. Explanatory correlates
could be investigated as competing factors, for example, in multiple regression
models,if linear associations are either present or can be effected. The latter
will require careful exploratory procedures to detect the appropriate transfor-
mation.
-------
32
It is possible that a linear combination of several responses would be more
effective in discriminating among groups than any single variate investigated
here. Statistical stepwise procedures for selecting "best" subsets from vari-
ables which show some difference across groups in these tables would suggest
appropriate vectors of variates.
In conclusion, although the univariate analysis employed here exhibited
limited capability to distinguish occupational groups for most responses, the
next step is to select combinations of responses which may prove to be more
effective. Significant correlations (between the laboratory variables them-
selves as well as between variables and concomitant factors such as exposure
to other possibly toxic agents) should be investigated. More detailed statis-
tical models would utilize the exhibited associations to explain the more than
normal variation that has been found in numerous responses in these preliminary
analyses.
2. Analyses Specific to PBB Exposure
The primary objective of the study was to determine the adverse health
effects related to PBB occupational exposure, as compared to other neighbor-
hood industrial workers. A valid and meaningful analysis should be from a
"dose-response" approach. However, in the absence of the PBB, PBBO, and PCB
serum levels, only general employment characteristics such as period and
length of employment were taken into consideration. Moreover, the univariate
approach to the analysis of the data considered the place of employment only
as a stratum, i.e., the exposure characteristics were limited to whether a
person had ever been or had not been exposed to PBB/P3B0, regardless of the
level of exposure. This approach was very limited as it completely ignored any
indication of the level of exposure. The P3Bs are highly lipotropic substances
which are .not easily metabolized or biodegradea; therefore, the current serum
level can represent the accumulated detectable exposure. Assessment of the
accumulated exposure, together with employment characteristics and manufacturing
data may lead to identifying high risk exposure periods. In addition, in the
analyses of health effects-related exposure, one should look not only for the
people with adverse health effects, but for trends and the association between
biological variates and level of exposures within a given "normal range".
Based on the observed serum, levels, we believe that the preliminary analysis
was unable to identify more adverse health effects if they existed because of
the heterogeneity of the FBB workers with respect to dose and duration. There-
fore, we suggest that the serurr, level of PBB/PBBO be examined as a function of
employment variables. We will then be able to identify a subgroup of workers
with a significant high level of exposure. The records of these people should
be examined to determine whether they have an excess of adverse health effects.
In addition to the simple dichotomy examined in the current study (i.e.,
PB3/not PBB employment), in order to complete the statistical analysis it will
be necessary to:
-------
33
1) define an index which describes a gradient of exposure to P3B. This
index of exposure would be based on a choice from the following multiple
parameters:
a) PBB serum levels
b) period of "active" exposure to PBB (i.e., length of employment at the
plant during period of actual PBB production)
c) duration of "passive" exposure (i.e., length of tine since last exposure).
2) examine the statistical association between the above index and the
medical variates. An unfavorable trend, even within the "normal" range of the
laboratory tests, may indicate more health problems in the future.
3. Epidemiological Considerations
The determination of whether occupational exposure to toxic substances such
as PBBs and P3B0s produced adverse health effects necessitates a carefully
planned analysis of the subsequent mortality or morbidity experience over time
of a group exuosed to the factor under consideration. In interpreting differences
in morbidity and mortality among occupational groups, it must be remembered that
1) state of health itself may determine entry into a specific occupation and 2)
the current co-worker group will not include those who have left employment due
to the deleterious occupational exposure although using the cohort approach may
have minimized some of these deficiencies.
An important limitation of the present study was the inadequate follow-up
time subsequent to PBB exposure. A longer period of observation could result
in identification of adverse health effects which would not be apparent in a
shorter time span. A similar study at a later date may detect latent adverse
health effects associated with PBB exposure.
SUMMARY
During the last week of August, 1978, a study of 36 men who had been
exposed to PBEs and PBBOs in their working environment for at least six weeks
between January, 1973 and August, 1978, was carried out. The plant had manu-
factured P33 and P^BO ur.t.i 1 production of the forxer was terminated in
April, 1977. Following cessation of the manufacture of decabromobyphenyl, the
factorv concentrated on producing decabromobypheny1 ether.
Controls were selected from workers employed in the same community without
exposure to PBB and PBBO with an atteir.pt to match the exposed population by aqe
and race.
In addition to demographic, smoking, drinking and lifetime occupation
questionnaires, self-reported medical histories, on-site general physical and
dermatological examinations, neurological ar.d nerve conduction studies and
laboratory tests cf urir.erblood , including PUB serun Levp". h, and nsasurps of
liver, thyroid and reproductive functions, were performed.
-------
34
Analysis of the data focused on:
1) Description of the univariate distributions of the laboratory variates.
Study ranges (95 percent) are reported in addition to analogous "normal" intervals
found in the literature.
2) Comparison of the distributions of the variables across various
selections of occupational groups, in particular the age, sex and race matched
triads, usinq the analysis of variance F-test for interval-scale measurement
and chi-square tests for classifications. For the nerve conduction velocities,
adjustment for age was made with an analysis of covariance model, where
assumptions were met.
Conclusions are:
1) As can be read in Table II, for the study as a whole, a large majority
of responses show either a wider than normal 95 percent ranae and/or a distri-
bution that is off-center, frequently in a direction implying health impairment.
For examole, all five nerve velocities studied are reduced, suggestive of neuro-
toxicity for the study as a whole.
2) As can be seen from Table V to IX, procedures which were used to test
for differences across occupational groups (the triads) verv rarely showed
significance. The several medically-isolated instances, e.g., serum phosphorus
(P = .03, Table VII), may well be chance associations.
However, for two categories of variates, an adverse health effect from
occupational PB3 exposure may be postulated. As reported on Page 19, this
study found a significantly increased incidence of hyoothvroidism amona the
*3 (r
PBR workers. Also, sural sensory velocities (discussed on Page 28) have
shown a significant reduction (P < .025) when compared with the electrical
wiremen. Following adjustment for age, peroneal motor velocities are reduced
(P < .05) when compared with the steel group. Pending further adjustment oro-
cedures for age, it is possible that the PBB median motor velocities also are
reduced. Further investigations in regard to both hypothvroidism and nerve
velocities, in particular, looking for possible dose-response associations
within the P3B grouo, are warranted.
-------
35
REFERENCES
1. Sanders, H.J. Flame retarrfants — special report. Cheir.. Eng. News. 22-36,
1978.
2. Matthews, H.B., S. Kato, N.K. Morales, D.B. Tuev. Distribution and excretion
of 2,4,5,2,4,5-hexabromobipheny1, the major component of Firemaster BP-6.
J. Toxicol. Environ. Health 3:599-605, 1977.
3. Munma, C.E., D.C. Wallace. Survey of industrial processing data. Task II -
Pollution potential of polybrominated biphenyls. EPA-560/3-75-004. June,
1975.
4. Kinbrough, R.D. The carcinogenic and other chronic effects of persistent
halogenated organic compounds. Health Effects of Halogenated Aromatic
Hydrocarbons. Annals of the New York Academy of Sciences 320:415-418,
1979.
5. Bahn, A.K. et al. Melanoma after exposure to PC3s. Letter to the Editor,
New. Eng. J. Med. 295:450, 1976.
6. Jackson, T.F., F.L. Halbert. A toxic syndrome associated with the feeding
of polybrominated'biphenyl contaminated protein concentrate to dairy
cattle. J. An. Vet. Assoc. 165:437-439, 1974.
7. Isleb, D. R, G.L. Whitehead. Polybrominated Biphenyls: An Agricultural
Incident and its Consequences I. The Agricultural Effects of Exposure.
Proceedings at the Ninth Annual Conference on Trace Element Substances in
Environmental Health, pp. 47-54, 1975.
B. Dent, J.G. et al. The induction of hepatic microsomal metabolism in rats
following acute administration of a mixture of polybrominated biphenyls.
Toxicol. Appl. Pharmacol. 38:237, 1976.
9. Lee, K P. et al. Octabroncbiphenvl-induced ultrastructural changes in rat
liver. Arch. Environ. Health 30:465, 1975.
10. Aftosir.is, J.G. et al. Toxicology of brominated biphenyls. I. Oral toxicity
and enbryotoxicity. Toxicol. Appl. Phamacol. 22:316, 1972.
11. Ringer, R. , D. Polir.. The Biological Effects of Polybrominated Biphenyls
in Avian Species. Federation Proceedings 36:1894-1998, 1977.
12. Allen, J.R. , D.A. Barsotti, U.K. Lambrecht, J.P. Van Miller. Reproductive
effects of halogenated aromatic hydrocarbons on nonhnman primated. Health
Effects of Halogenated Aromatic Hydrocarbons. Annals of the New York
Academy of Sciences 320:419-425, 1979.
-------
36
13. Kimbrouah, R.D., V.W. 3urse, J.A. Liddle, Persistent liver lesions in rats
after a single oral dose of polybrominated biphenyls (Fire Master FF-1)
and concomitant PBB tissue levels. Environ. Health Perspect. 23:265-273,
1978.
14. Selikoff, l.J. et al. Health effects of exposure to polybrorr.inated biphenyls.
Results of a clinical field survey. Nov. 4-10, 1976. Interim summary
report to National Institutes of Environmental Health Sciences. ES 00928.
15. Meester, W.D., D.J. McCoy. Human toxicology of polybrominated biphenyls.
Clin. Toxicol. 10:474, 1977.
16. Budd, M.L. et al. Polybrominated Biphenvl Exposure - Michigan, Morb. and
Mort. Weekly Report 27:115-121, 1978.
17. Anderson, H.A. et al. Symptoms and clinical abnormalities following ingestion
o~ polybrominated-biphenyl-contaminated food products. Health Effects of
Halogenated Aromatic Hydrocarbons. Annals of the New York Academy of
Sciences 320:684-702, 1979.
18.. Landrigan, P.J. et al. Cohort study of Michigan residents exposed to poly-
brominated biphenyls: epidemiologic and immunologic findings. Health
Effects of Halogenated Aromatic Hydrocarbons. Annals of the New York
Academy of Sciences 320:284-294, 1979.
19. Bekesi, J.G., J.F. Holland, H.A. Anderson, A.S. Fischbein, w. Rom, M.S.
Wolff, I.J. Selikoff. Lymphocyte function of Michigan dairy farmers ex-
posed to polybrominated biphenyls. Science 199:1207-1209, 1978.
20. DeCarol, V.J. Studies on bromir.ated chemicals in the environment. Health
Effects of Halogenated Aromatic Hydrocarbons. Anr.als of the New York
Academy of Sciences 320:678-681, 1979.
21. Selikoff, I.J. Personal communication. June, 1978.
22. Ancr.. Is PBB a human health hazard? Mich. Med. 75:560, 1976.
23. Isbister, J.L. The short tern ef fects of PBB on health. r-lich. Uect. o-"
Public Health report. Feb. 23, 1977.
24. Syster, J.T. Breast milk studies. Mich. Depart, of Public Health.
Nov. 18, 1976.
25. Kay, K. Polybrominated biphenyls (P3B) environmental contamination in
Michigan 1973-1976. Environmental Res. 13:74, 1977.
26. Valciukas, J.A. et al. Polybrominated biphenyls: neuroioqical and
behavioral effects. Report to NIEHS, May 26, 1978.
-------
37
27. Anderson, H.A. Prevalence of unanticipated symptoms among dairy farmers
in Michigan and Wisconsin. In Health Effects of Exposure to PBB. Interim
summary report to NIEHS, 1976.
28. Valciukas, J.A. et al. The neurotoxicity of polybrominated biphenyls:
results of a medical field study. Health Effects of Halogenated Aromatic
Hydrocarbons. Annals of the New York Academy of Sciences 320:3 37-367,
1979.
29. Chiu, E. PBB: An American Tragedy. Prentice-Hall: Enqlewood Cliffs,
N.J., 1979.
30. Housworth, J. Chronic Diseases Division, Center for Disease Control,
Atlanta, personal communication, September 27, 1977.
31. Hunter, J. et al. Increased hepatic microsomal enzyme activity from
occupational exposure to certain organochlorine pesticides. Nature 237:
399-401, 1972.
32. Bahn, A.K., O. Bialik. Internal report to NIOSH, April 7, 1978.
33. Dixon, W.I., M.B. Brown. Biomedical Computer Programs P-Series, University
of California Press, Berkeley, 1977.
34. Kilk, M.B., R. Gnandesikan. Probability plotting methods for the analysis
of data. Biometrika 55:1-17, 1968.
35. Tukey, J. Exploratory Data Analysis. Addison-Wesley. Reading, Mass., 1977.
36. Bahn, A.K. et al. Hypothyroidism in workers exposed to polybrominated
biphenyls. N. Enc. J. of Med. 302:31-33, 1980.
37. Rowlands, C., V.D. Castracane, J. Seifter- Effects of polybrominated
biphenyls (PBBs) on Thyroid, Adrenal and Testis Function in the Rat.
61st Annual Meeting of the Endocrine Society, Abstract 913, 1979.
-------
TABLE I
Hematologic, Serologic, Urinary and Electrodiagnostic Tests, Methods of Collection, and Testing Laboratory
BLOOi) TASTING S?EC1ML:Q LArf(...-ATCR\ TESTS*
1.
TEST
A. CBC, Differential, hemoglobin and
and hematocrit
SMA 24 (glucose, BUJ, creatinine,
Na, CI, K, C02 as bicarbonate,
uric acid, Ca, P, total protein,
albumin, globulin (total),
alkaline phosphatase, bilirubin
indirect and direct, cholesterol,
SCOT, SGl'T, LDH, triglycerides,
ratio BUN/creatinine, ratio
albumin/globulin) plus
YGT.
Resin T^ Uptake
B. Testosterone, assay
FSH, LH, and TSH assay, T^
Triiodothyronine T^ (if hyper-
thyroidism suggested by T^ and
TSH values), thyroid antibodies
(if hypothyroidism indicated by
T. and TSH)
All of above by radioimmunoassay
except thyroid antibodies
C. PBB, PBBO, PCB levels in serum
METHOD OF COLLECTION
(1) 7 ml. purple top (EDTA) tubes
must be examined within 2U hrs.
(3) 10 ml. red top tubes (Corvac)
Allow to stand at room temperature
for 30-60 minutes for clot
retraction. Then centrifuge at
6000 RPM for 10-15 minutes.
Refrigerate immediately.
(2) 10 ml. red top tubes (Corvac)
Handle as above
Freeze wi thin 12 hours
(2) 10 ml. red top (Corvac)
Obtain serum from whole blood as
above. Aspirate serum into glass
scintillation vial and refrigerate
immediatelv. Freeze, within 12 hrs.
TESTING LABORATORY
Dr. Leslie M. Shaw, Director
Toxicology Laboratory, Division
of Laboratory Medicine, Path^lo;
Department, HUP
Bioscience Laboratories
Dr. Peter Snyder, Endocrine
Section, Dept. of Medicine,
U of P School of Medicine
Remaining serum kept frozen
indefinitely
Bioscience Laboratories
Research Triangle Institute
North Carolina
Dr. Etio Pellizzari
*10a of all samples retested.
All specimens were brought back to the respective laboratories in Philadelphia at the en-1 -T; each day by
courier from Bioscience Laboratories.
-------
TABLE I (continued)
UniNE LABORATORV TEfTS*
11. TEST
SPECIFICATIONS
RESPONSIBILITY FOR LAB WORK
A. Ratio of the following parameters
to creatinine.
1. porphobilinogen
2. delta-aminolevulinic acid
3. coproporphyrins total
4. uroporphyrins \ porphyrins
Creatinine mg/ml
Pour urine into Bioscience plastic
vial A1 for assay of 1, 2, and 3,
vial A8 for assay of 3 and 4.
Refrigerate immediately. Freeze Al
within 12 hours. Protect from
light.
Dr. Carl Alper, Bioscience Laborator
B. Ratio glucaric acid to creatinine.
Pour urine into container provided by
U of P laboratory. Refrigerate
immediately.
Dr. Leslie M. Shaw, Pepper Lab, HUP
Ihe urine tests weri' done on both a spot sample and a 24-hour sample when available. To standardize
the collection of spot urines, screenees were asked to void 2 hours before entry. Upon entry, and over the
3 hour period at the exam site, the screenee collected his urine in a 500 cc. flask. Only sugar-free sodn
and water were offered to screenees until all samples were collected.
The 24 hour urines were collected in % gallon jugs (Bioscience Lab) wrapped in pap-?r bags. Preservatives
or refrigeration were not required.
Total volume of both spot and 24-hour urines were measured and then aliquots fron each handled identically ac-
indicated above.
QUALITATIVE URINE TESTS
Dr. Shaw's research assistant did qualitative urine dipstick on each specimen on rice at the laboratory to
ketone, nitrites.
determine pH, glucose, blood, protein,
W
-------
ELECTRODIAGNOSIS
TABLE I (continued)
III.
TEST
METHODS
RESPONSIBILITY FOR LAB WORK
A. Sensory nerve conduction
arm: median
ulnar
leg: sural
TECA Electromyogram & Averager
1) stimulate nerve distally by ring
electrodes on fingers and surface on
foot. 2) proximal recording (NAP)
by surface electrodes at one or more
points; orthodromic
Prof. R.E. Lovelace, Director of EMG &
Clinical Neurophysiology Laboratory,
Neurological Institute, NY, 710 W. 168th
St., NY, NY 10032 (& colleagues including
Ms. Betty Hughley).
B. Motor nerve conduction
arm: median
leg: peroneal
TECA Electromyogram & Averager
1) stimulate nerve at one or more
proximal points and one distal.
2) record distal MAP orthodromically
by surface electrode over motor point.
C. F-Waves
leg: peroneal
TECA Electromyogram & Averager
"Rebound" late response on supramaximal
motor stimulation as above (compute from
L1/T12). In A, B & C above exact
temperature recording control and monit
oring essential.
J?
O
-------
TABLE II
Normal Range a nd Study Results of Test Va r iabl c;s
Normal Range
STUDY RESULTS
Mean ^ Std._ Uev. 95% of Study
A. Nerve conduction
1. Median motor
2. Peroneal motor
3. Median sensory
4. Ulnar sensory
5. Sural sensory*
B. Hematology
>50.3
>45.6
>50.9
>42.7
>40.3
1.
2.
3.
6.
7.
8.
9.
10.
11.
12.
13.
Hemoglobin
(gms %)
Hematocri t
(%)
Red blood cells
(RBC) (millions/mm^)
Mean Corpuscular Vol.
(MCV) (micron^)
Mean Corpuscular Hemglobin
(MCH) y/y
Mean Corpuscular
Hemoglobin Concentration
MCI1C (%)
Leucocytes (1000/mm^)
Neutrophils (%)
Band (%)
Eosinophil* (%)
Basophil* (%)
Monocyte (%)
Lymphocyte (%)
14-18
12-16
38-54
36-47
4.5-6.0
4.3-5.5
79-99
27-35 '
22-37
5-10
40-60
0-5
1-3
0-1
4-8
20-40
57 .291
46.206
54.366
50.585
46.750
15.388
43.630
4.896
89.159
31.558
35.202
7.463
53.195
4.775
2.036
0.493
6.674
32.905
5.129
4.399
6.371
5.842
6.047
1.037
3.090
0.412
5.019
1.929
0.968
1.960
9.935
3.477
1.757
0.776
3.344
8.845
>50.0
>39.8
>44.8
>41.1
>37.1
M 13.7-17.5
Ft 13.3-15.2
M 38.5-49
Ft 37.5-44
M 4.1-5.5
FY 4.1-5.0
81-97
28.5-34.5
33.9-36.6
4.5-11.5
37-71
0.5-12.5
0.2-5.5
0-2.9
1.5-13.5
16.5-49.5
* A logarithmic transformation was made for these variables (see Paf»e 13)
+ For 8 females -- 100% range of the study
-------
TABLE II (continued)
Normal Range and Study Results of Test Variables
C. Blood Chemistry
1.
Sodium
(meq/L)
2.
Potassium
(meq/L)
3.
Chloride
(meq/L)
4.
Carbon dioxide
(C02) (mM/I.)
5.
Blood urea nitrogen
(BUN) (mg/dl)
6.
Creatinine
(mg/dl)
7.
Glucose*
(mg/dl)
8.
Calcium
(mg/dl)
9.
Phosphorus
(mg/dl)
10.
Total protein
(g/dl)
11.
Uric acid
(mg/dl)
12.
Acid phosphatase
(U/L @ 30° C)
13.
Cholesterol
(mg/dl)
14.
Total bilirubin
(mg/dl)
Normal Range STUDY RESULTS
Mean Std. Dev. 95% of Study
138-146 141.094 2.746 136-145
3.5-5.3 4.195 0.357 3.7-5.0
97-108 102.401 3.083 96-106
24-32 24.043 1.981 21-27
10-20 15.664 4.138 8-25
M 0.8-1.4 M 0.9-1.4
F 0.7-1.2 1"115 °-163 Ft 1.0-1.0
60-105 85.875 22.556 68-113
8.5-10.5 10.027 0.430 9.4-10.9
2.4-4.4 3.488 0.597 2.3-4.7
5.8-7.4 7.871 0.442 6.9-8.6
M 3.5-8.0 M 4.5-8.5
F 3.0-6.0 5.648 1.388 p+ 2.9_4.6
1.7-5.0 2.717 0.533 1.5-3.5
190.416 31.744 135-255
see note
following 1) 23
0.2-1.2 0.425 0.483 0.1-1.1
-------
TABLE II (continued)
Normal Range and Study Results of Test Variables
Normal Range
Mean
STUDY RESULTS
Std. Dev.
-9JL% of Study.
C . Blood Chemistry
(continued)
15. Aspartate Amino
Transferase*
(SGPT (U/L @ 30°C)
M 9-38
F 6-25
23.625
16.495
16. Gamma Glutamyl
Transferase* (yGT)
(U/L @ 30°C)
M 6-40
F 4-24
33.153
46.735
17. Lactic acid dehydrogeno
(LDH) (U/L @ 30°C)
Se 140-270
281.116
77.174
18. Albumin
(g/dl)
3.6-4.8
5.031
0.334
19. Globulin
(g/dl)
1.6-3.5
2.838
0.354
20. Albumin globulin ratio
1.1-2.2
1.812
0.273
21. Alanine Amino
Transferase* (SGOT)
(U/I. @ 30°C)
M 10-25
F 9-22
23.625
16.495
22. Alkaline phosphatase
(U/L)
12-40
24.941
9.211
23. Triglyceride'"
(mm ol/L)
see note
following #23
1.993
1.592
M 11-81
Ft 10.5-477
M 5-130
Ft 9-67
195-555
4.4-5.6
2.3-3.6
1.4-2.3
M 10-81
Ft 9-15
16-39
1.02-6.751
Note
0-19 years
20-29 years
30-39 years
40-49 years
50-59 years
CHOLESTEROL2
Normal Ranges
<190 mg/dl
<220 mg/dl
<240 mg/dl
<270 mg/dl
<290 mg/dl
<1.58 mmo1/L
<1.58 mmol/L
<1.70 mmol/L
<1.81 mmol/L
<2.15 mmol/L
TRIGLYCERIDES2
Normal Ranges
(140 mg/dl)
(140 mg/dl)
(150 mp,/dl)
(160 mg/dl)
(190 mg/dl)
High levels in some individuals probably due to recent meal
Constant diet and 12-16 hour fasting necessary prior to sampling
-------
TABLE II (continued)
Normal Range, and Study Results of Test Variables
Normal Range
Mean
STUDY RESULTS
Std. Dev.
_.95% of Study
D. Endocrine
1.
2.
5.
6.
7.
Testosterone.
(mg/dl)
Follicle Stimulating
Hormone* (FSH)
(mlU/ml)
Luteinizing Hormone""
(LII) (mlU/ml)
Thyroxine (Tl,)
^ig/dl)
Thyroid Stimulating
Hormone* (TSH)
^yU/tnl)
Resin Uptake Ratio (RUR)
Free Thyroxine Index (FTI)
300-1200
<14
<18
A.5-11.5
<8
0.85-1.17
3.8-10.8
E. Urine
Spot:
1. Creatinine (mg/1)
2. De l t.a-Aminol itvulinic
Acid (mg/l/mg creat.)
3. Porphobilinogen quant.
(M/l/mg creat.)
4. Corphoporphyrin
(M/l/mg creat.)
5. Uroporphyrin
(M/l/mg creat.)
6. Gluaric acid
(U moles/l/mg creat.)
648.182
8.466
11.723
10.283
5.161
1.005
10.349
969.335
2.136
0.751
89.153
9.978
2 25.062
14.596
15.186
13.604
15.521
0.776
15.663
676.088
1.320
0.583
61.800
9.848
330-1050
1.2-15
1.6 (99%)
7.0 (99%)
7.7 (99%)
1.0 (99%)
7.02 (99%)
275-2850
1.2-5.5
0.19-2.03
34-210
1.0-28
10-43
-------
TABLE II (continued)
Normal Range and Study Results of Test Variables
Normal Range STUDY RESULTS
_ Mean Std. Dev. 95% of Study
F . Urine
(continued)
24-Hour Urine:
7. Creatinine (mg/24 hr)
1000-16000
1134.525
562.416
375-2475
8. Delta-aminolevulnic
acid (mg/24 hr)
1.3-7.0
2.615
1.209
0.90-5.4
9. Porphobilinogen
(ug/24 hr)
0
1
ro
00
0.681
0.447
0.08-2.0
10. Corphoporhyrin (ug/24 hr)
50-250
81.632
60.843
15-165
11. Uroporphyrin (ug/24 hr)
10-50
13.040
8.004
4.5-31.5
12. Glucaric acid
(m moles/24 hr)
4-74
10-35
-------
n?,
TABLE III
THE 36 STUDY TRIADS (MALES ONLY)
--
!""" pbb"
Workers
- • - -
riciar.s
& Others
Steel
Workers &
O t h e
f r i a d
Birth
3 i r t h
Apft
Birth
Aye
N u mbc-r
Nunber
Year
Number
Year
Diff . 1
Number
Year
D i f c
002
1956
00 7 2
19 56
0
005
1956
0
2
009
195 5
00 6 2
1956
+ 1
004
1956
+ 1
3
008
1955
013 2
1954
-1
003
1956
+ 1
4
012
1954
01 7 2
1954
0
011
1955
-^1
5
025
1953
01 6 2
1954
+ 1
015
1954
+ 1
6
0 24
1 953
0 2 7 2
1953
0
014
1954
+ 1
7
023
1953
| 0 3 4 2
19 52
-1
013
1954
+ 1
8
022
1953
0 4 3 2
1951
-2
026
1953
0
9
021
19 5 3
029
1952
-1
032
1952
-1
10
020
1953
030
1952
-1
031
1952
-1
11
028
19 5 2
0 39
1951
-1
04 0
1951
- ]
12
0 36
1951
038
1951
0
041
1951
0
J 3
044
19 50
037
1951
+ 1
0 4 6 2
1950
0
i 4
04 7
1949
045
1950
+ 1
050
19 4 9
0
15
054
1948
049
1949
+ 1
059
1948
0
1 3
053
1948
05 6
1 948
0
058
1948
0
17
05 2
1948
055
1948
0
057
1948
0
1 8
051
1948
06 2
1947
-1
0 6 3 2
1947
-1
1 9
060
1947
06 1
194 7
0
0 6 6 2
1946
-1
20
071
1944
067
19 4 j
+ 1
070
] 945
•n
2 1
083
1938
075
19 4 0
+ 2
074
1941
+ 3
22
0 84 3
1938
07 3
1942
J-4
0823
1939
+ 1
¦ 3
089
193 7
081
19 39
+ 2
0 7 8 2
1940
+ 3
24
088
1937
080
1939
+ 2
0 7 7 2
1940
+ 3
:>5
087
19 3 7
08 5
1938
+ 1
0 7 6 2
1940
+ 3
26
092
1936
093
1936
0
08 6 2
1938
+ 2
? 7
091
1936
09 7
19 35
-1
09 02
1937
+ 1
..8
095
1935
101
1931
-4
102
1931
-L
29
094
1935
100
1931
- L
108
1930
-5
>0
111
1928
113
1928
0
106
1930
+ 2
31
112 3
:. 9 2 3
114
1928
0
109 3
1930
+ 2
2
119
"9 2 5
122
.1 9 2
-1
1 10
19 29
+ 4
3 3
123
19 2 3
121
1924
+1
116 2
1928
+ 5
34
12 4
19 2 3
127
1923
0
1 312
1921
_ 2
.15
123
1 923
126
1923
0
1 3 3 2
1920
-3
36 i
129
1921
130
1921
0
135
1918
-3
Mean
42.3
42.3
42.7
S-D. ,
11.15 |
11.18
_
1_1_. 2_2._
Note: I.D. number recoded for matching purposes.
^ge Diff. is age difference in years between the PBB worker and his matched
case .
?Subject from residual control category.
Member of a minority race.
-------
TABLE IV
Preliminary Statistical Analysis of Laboratory Tests*'
a. ANOVA among tested groups
Grouping
Vsriable
analys is
Test
for
normality
(triads)
3 3'
(employment)
4 4'
(PBB workers
vs. controls)
(30 groups)
Urine X
Endocrine X
Hematology X
ciloo'i Chemistry X
Nerve Conduction X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
b. Analysis of covariance (age)
Nerve Conduction - - - X
X
. Correlation of laboratory data with age
On original variables
On transformed variables
Urine
Endorcrine X
Hematology X
Blood Chemistry
Norve Conduction X
Nerve Conduction
(radical exclusions) X
X
X
X
X
i. Correlation of laboratory data with length of employment at
PBB production plant
Ji ine
Endocrine X
Hematology X
Blood Chemistry X
X
X indicates analyses was oer^ormed
- indicates analyses not considered appropriate or necessary
J'l>ilizing BMDP Biomedical Computer Programs (Dixon, 1977)
-------
48
TABLE V
Reported Symptoms by Occupational Group
System
PBB Workers
(n=39)
Number
of persons
%
Electricians
(n=42)
Number
of persons %
Steel Workers
(n-26)
Number
of persons %
General, e.g.,
weight loss
fatigue
15
38.5
13
31.0
9
34.6
Eyes, ears
nose, throat
14
35.9
17
40.5
12
46.2
Respiratory
13
33.3
15
35.7
10
38.5
Gastrointestinal
10
25.6
9
21.4
10
38.5
Genitourinary
12
30.8
17
40.5
8
30.1
Endocrine
19
48.7
19
45.2
12
46.2
Dermatologic
27
69.2
32
76.2
17
65.4
Musculoskeletal
12
30.8
17
40.5
8
30.1
Neurologic
22
56.4
26
61.9
14
53.4
No significant difference was found
between
the three
study groups
(Chi-square
test).
-------
TABLE VI
Number of Workers with Abnormal Neurological Findings
v Group PBB
Neurologic aspect (39)
Electricians
(42)
Steel Workers Others
(26) (28)
Cranial nerves
Pathological reflex
Deep tendon reflex
Motor and coordination
Gait and stance
T remor
Sensory; pin prick
Mini-nental State
(•124 points)
9
2
2
No significant difference is found in comparison of the study groups
(chi-square test).
-------
TABLE VII
Comparison of the Mean Response of Laboratory Results Among the Three Study
Matched Triads
Laboratory
Test
Mean Response
PBB Workers
(n=36)
Electricians & Steel Workers & Total
Others (n=36) Others (n-36) (n=10S)
Analysis of
Variance -
F statistic
P
Value
Hematology
Hemoglobin
Hematocrit
RRC
MCV
MCH
MCHC
Leucocytes
Neutrophils
Band
Eosinophils''
Basoph iIs*
Monocytes
Lymphocytes
15.60
44.08
4.92
89. 70
31.81
35.40
7.18
53.11
5.03
2.14
1.53
6.31
33.60
15 .30
43.40
4.90
88.60
31.30
35.10
7.48
54.80
4.82
2.59
1.40
7.11
31.33
15.61
44.14
5.00
89. 10
31.60
35.30
7 .44
49.64
6.41
2.80
1. 30
6.81
34 .83
15.50
43.90
A .93
89.10
31.60
35.23
7.37
52.51
5.44
2.52
1.42
6.74
33.24
1 .356
0.679
0.248
0.396
0.630
1.068
0.246
2.584
2. 328
1.088
0.357
0.567
1.293
0.26
0.51
0. 78
0.67
0.53
0. 35
0.78
0.08
0.10
0. 34
0.70
0.57
0.28
-------
TABLE VII (continued)
Comparison of the Mean Response of Laboratory Results Among the Three Study Matched Triads
Laboratory
Test
PBB Workers
(n=36)
Mean Response
Electricians & Steel Workers & Total
Others (n=36) Others (n=36) (n=lC3 )
Analysis of
Variance -
F statistic
P
Value
Blood Chemistry
Sodium 141.50
Potassium 4.30
Chloride 102.81
CD2 24.11
Blood Urea
Nitrogen 15.69
Creatinine 1.13
Glucose 87.08
Calcium 10.08
Phosphorus 3.32
Total Protein 7.95
Uric Acid 5.71
Acid Phosphatase 2.83
Cholesterol 188.60
140.40
4.13
101 .64
24 .03
15 ,94
1 .14
86. 14
10.20
3.70
7.90
5.83
2.71
191.90
141 .30
4. 23
102.61
24 .08
15. 20
1.11
82.81
10.00
3.44
7.90
6.04
2.81
187.40
141 .10
4. 21
102.35
24.07
15.60
1.13
85.34
10.10
3.50
7.90
5.90
2.80
189.30
1.416
1 . 337
1.419
0.018
0.362
0.263
0.427
1.501
3.491
0.423
0.522
0.575
0.175
0.25
0.27
0.25
0.98
0.70
0.77
0.65
0.23
0.03
0.66
0.59
0.56
0.34
-------
TABLE VII (continued)
Comparison of the Mean Response of Laboratory Results Among the Three Study Matched Triads
Laboratory
Test
Mean Response
PBB
(n=36)
Electricians &
Others (n=36)
Steel Workers & Others
Others (n=36) (n=30)
Analysis of P
Total Variance - Value
(n-138) F statistic
Endocrine
Testosterone
FSH
log FSH
LH
log LH
T4
TSH
log TSH
RUR
FT I
653.3
5.83
0.66
10.12
0.88
7.72
6.08
0.41
0.94
7.20
613.3
6.19
0.71
8.81
0.89
8.45
1.87
0.24
0.94
7.80
708.4
6.03
0.72
9.46
0.90
8.10
2.07
0.27
1.0
7.63
7.97
0.73
10.84
0.89
2.04
0.26
659.3
6.43
0.70
9.76
0.89
8.10
3.06
0.30
0.94
7.52
1.590
1.122
0.332
0.417
0.353
0.861
3.755
2.566
0.588
0.880
0.21
0.34
0.80
0.74
0.99
0.43
0.01
0.06
0.56
0.42
r1
-------
TABLE VII (continued)
Comparison of the Mean Response of Laboratory Results Among the Three Matched Study Triads
Laboratory
Test
PBB Workers
(n=36)
Mean Response
Urine
Spot Urine
Creatinine 1058.61
Aaminolevuli n Lc
acid 2.22
Porphobilinogen 0.80
Coproporphyrin 96.50
Uroporphyrin 10.10
Electricians & Steel Workers & Total
Others (n=36') Othpr.s Cn=3(S') (n=108)
1034.44
2.40
0.90
95:4 2
9.30
993.61
2.20
0.80
87.80
10.92
1028.89
2. 30
0.80
93.22
10.10
Analysis of
Variance -
F statistic
P
Value
0.076
0. 150
0.331
0.180
0.427
0.93
0.86
0.72
0.83
0.65
24-hour Urine (27 cases)
Creatinine 1157.20
^ aminolevulinic
acid 2.64
Porphobilinogen 0.70
Corproporphyrin 68.04
Uroporphyrin 11.50
(11 cases)
1368.64
3.00
0.92
76.10
14.10
(17 cases)
1114.40
2.80
0. 72
81 .41
13.90
(55 cases)
1186.24
2.75
0.73
73.80
12.74
0.681
0.272
1.108
0.480
0.977
0.51
0.76
0.34
0.62
0.38
*See log-response for better comparison.
-------
TABLF, vil (continued)
Comparison of the Mean Response of Laboratory Results Among the Three Study Matched Triads
Laboratory
Test
PBB Workers
(n=36)
Mean Response
Electricians & Steel Workers & Total
Others (n=36) Others (n=36) (n- 108)
Analysis of
Variance -
F statistic
P
Value
Blood Chemistry
(cont.)
Total Bilirubin 0.50
SGPT 20.30
yGT 32.72
LDH 285.80
Albumin 5.14
Globulin 2.81
Albumin Globulin
Ratio 1.90
SCOT 2 2.94
Alkaline
Phosphatase 24.08
Triglycerides 2.08
0.34
23.81
41 .36
283.22
5:10
2. 82
1.82
27.00
27.20
2.30
0.40
21.70
36. 14
282.14
5.00
2.90
1 .81
26.64
24 .20
1 .80
0.40
21 .92
36.74
283.70
5.07
2.83
1 .83
25.52
25. 16
2.05
1 .676
0.877
0.260
0.019
2.076
0. 282
0.486
0.570
1 .150
0.977
0. 19
0.42
0.77
0.98
0.13
0.75
0.62
0.57
0.32
0.38
-------
TABLE VII(Cont)
Comparison of the Mean Response of Laboratory Results Among the Three Study Matched Triads
Mean Response
Nerve Conduction PBB Electricians Steel Workers Total Analysis of P
Velocity* Workers and Others and Others Variance - Value
F statistic
Median Motor
56.78
(32)
56.45
(28)
59.01
(33)
57.47
(93)
2.02
.1381
Peroneal Motor
45.36
(32)
44.80
(29)
47.43
(32)
45.90
(93)
3.39
.0381
Median Sensory
55.09
(34)
54.33
(29)
53.87
(33)
54.44
(96) '
.421
.6574
Ulnar Sensory
51.04
(33)
50.78
(27)
50.14
(28)
50.68
(88)
.186
.8306
Sural Sensory
43.96
(33)
48.77
(30)
45.88
(33)
46.12
(96)
4.68
.0116
Log Sural Sensory
1.64
(33)
1.685
(30)
1.657
(33)
1.66
(96)
4.95
.0090
*The triads are not age-matched (note reduced number of subjects in ()) due to velocities not available, e.q., subject
requested no testing. This table is presented in the interest of completeness, i.e., to offer a comparison across
study triads on nerve conduction velocities. The reader is referred to Page *7 for the results of comparing across
"pure" occupational grouns, after subjects with all known extraneous causes for neurotoxicity have been excluded,
and age-adjustment is attempted via analysis of covariance.
VT\
-------
TABLE VIIT
Percentage Distribution of
Laboratory Results:
Below, Within pnd
Above Normal
Range
High
Chi
Square
Sig.
Level
Laboratory
Test
Group
Comparison
of 108
Persons
Tnclu
ded in the
Age-Matched
Triads
Normal
Low
Levels
Normal
High
Levels
Low
PBB
E
S
PBB
E
S
PBB
E
S
Total 108 Persons
Hematology
Hgb.
5.6
11.1
2.8
94.4
88.9
97.2
-
-
-
6.5
93.5
-
2.13
0.34
Hct.
-
-
-
100.0
100.0
100.0
-
-
-
-
100 .0
-
-
-
RBC
19.4
8.3
13.9
80.6
88.9
86.1
-
2.8
-
13.9
85.2
0.9
3.75
0.44
MCV
2.8
2.8
2.8
94 .4
97.2
07.2
2.8
-
-
2.8
96.3
0.9
2.01
0.73
MCH
-
2.8
2.8
97. 2
94 .4
97.2
2.8
2.8
-
1.9
96.3
1.9
2.01
0.73
MCHC
-
-
-
97.2
100.0
97.2
2.8
-
2.8
-
98.1
1.9
1.01
0.60
Leucocyte
11.1
2.8
2.8
80.6
88.9
86.1
8.3
8.3
11.1
5.6
85.2
9.3
3.35
0.50
Neutrophi1
8.3
11.1
11.1
69.4
58. 3
77.8
22.2
30.6
11.1
10.2
68.5
21.3
4 .39
0.35
Band
-
-
-
66.7
69.4
50.0
33.3
30.6
50.0
-
62.0
38.0
3.38
0.18
Eosinophil
19.4
11.1
11.1
80.6
83. 3
80.6
-
5.6
8.3
13.9
,81-5
4.6
4 .02
0.40
Basophi1
-
-
-
86.1
86.1
91.7
13.9
13.9
8.3
-
88.0
12.0
0.69
0.70
Monocyte
16.7
13.9
19.4
69.4
52.8
52.8
13.9
33.3
27.8
16.7
58.3
25.0
4.36
0.35
Lymphocyte
8.3
11.1
-
69.4
75.0
77.8
22.2
13.9
22. 2
6.5
74.1
19.4
4.74
0.31
Key -Co Abbfidv-icitA-OM'• PBB, Polybrominated Biphenyl Workers; E, Electricians & Others; S, Steel Workers & Others;
NA, Not Applicable; No Cases.
-------
TABLE VIII (continued)
Percentage Distribution of Laboratory Results: Below, Within and Above Normal Range
Group Comparison of 108 Persons Included in the Age-Matched Triads
Laboratory
Low
Levels
Normal
High Levels
Low
Normal
High
Chi
Sig.
Test
PBB
F.
S
PBB
E
S
PBB
E
S
Total 108 Persons
Square
Level
Blood Chemistry
Sodium
-
13.9
2.8
100.0
83.3
97.2
-
2.8
-
5.6
93.5
0.9
9.61
0.04
Potassium
-
-
-
100.0
100.0
100.0
-
-
-
-
100.0
-
-
-
Chloride
2.8
5.6
-
97.2
94.4
100.0
-
-
-
2.8
97. 2
-
2.05
0.35
CO 2
44.4
41 .7
33.3
55.6
58.3
66.7
-
-
-
39.8
60.2
-
1.00
0.60
Blood Urea
N itrogen
2.8
8.3
-
80.6
77.8
91.7
16.7
13.9
8.3
3.7
83.3
13.0
4.96
0.29
Creatinine
-
-
-
94.4
97.2
100.0
5.6
2.8
-
-
97.2
2.8
2.05
1.35
Glucose
-
2.8
-
94.4
94 .4
100.0
5.6
2.8
-
0.9
96. 3
2.8
4 .07
0.39
Calci um
-
-
-
91.7
86.1
91 .7
8.3
13.9
8.3
-
89.8
10.2
0.80
0.66
Phosphorus
2.8
-
5.6
07.2
83.3
91 .7
-
16.7
2.8
2.8
90.7
6.5
11.
) .02
Total Protein
-
-
-
8.3
25.0
13.9
91.7
75.0
86.1
-
15.7
84.3
3.90
0.14
[trie Acid
-
-
-
88.9
97.2
91 .7
11.1
2.8
8.3
-
92.6
7.4
1.89
0.38
Acid Phosphatase
-
5.6
-
100.0
94.4
100.0
-
-
-
1.9
-
-
-
-
Cholesterol
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-------
TABLE VIII (continued)
Percentage Distribution of the Laboratory Results: Below, Within and Above Normal Range
Group Comparison of 108 Persons Included in the Age-Matched Triads
Laboratory Low
Levels
Normal
High
Levels
Low
Normal
High
Chi
Sig.
Test FBB
E S
PBB
E
S
PBB
E
S
Total
108 Persons
Square
Level
Blood Chemistry (cont.)
Total Bilirubin 2.8
22.2 11.1
94 .4
77.8
88.9
2.8
12.0
87.0
0.9
8.29
0.81
SGPT _
_ _
83.3
75.0
88.9
16.7
25.0
11.1
-
82.4
17.6
2.43
0.30
yCT
-
86.1
77 .8
83.3
13.9
22.2
16.7
-
82.4
17.6
0.89
0.64
LDH
-
52.8
50.0
52.8
47.2
50.0
47. 2
-
51 .9
48.1
0.07
0.96
Albumin
-
13.9
30.6
19.4
86.1
69.4
80.6
-
21.3
78.7
3.09
0.21
Globulin
-
97.2
100.0
91.7
2.8
-
8.3
-
96. 3
3.7
3.63
0.16
Albumin Globulin
Ratio
-
-
-
-
-
-
-
-
-
-
-
SGOT
-
88.9
83.3
86.1
11.1
16.7
13.9
-
86.1
13.9
0.46
0. 79
Alkaline
Phosphatase
-
100.0
97.2
100.0
-
2.8
-
-
99.1
0.9
2.02
0.36
Triglycerides
-------
TABLE VIII (continued)
Percentage Distribution of the Laboratory Results: Below, Within and Above the Normal Range
Group Comparison of 108 Persons Included in the Age-Matched Triads
Laboratory Low Levels Normal High Levels Low Normal Hieh Chi Sig.
Test PBB E S_ PBB If S PBB E S Total 108 Persons Square Level
Endocrine
Testosterone
-
-
-
100.0
100.0
97.2
-
-
2.8
-
99.1
0.9
1.96
0.37
FSH
NA
NA
NA
91 .7
94.4
100.0
8.3
5.6
-
NA
95.4
4.6
2.93
0.23
LH
NA
NA
NA
86.1
94.4
88.9
13.9
5.6
11.1
NA
89.8
10.2
1 .41
0.49
Ti*
8.3
5.6
5.6
77.8
91 .7
88.9
13.9
2.8
5.6
6.5
86. 1
7.4
3.98
0.40
TSH
NA
NA
NA
86.1
97.2
100.0
13.9
2.8
-
NA
94.4
5.6
7.41
0.02
RUR
8.3
16.7
13.9
91.7
80.6
86.1
-
2.8
-
13.0
86.1
0.9
3.25
0.51
FTI
8.3
2.8
-
83.3
91.7
94 .4
8.3
5.6
5.6
3.7
89 .8
6.5
4 .05
0.39
Urine-24 hour
D-Amino Acid
7.4
9.1
5.9
92.6
90.9
88.2
-
-
5.9
7.3
90.9
1.8
2.35
0.67
Porphobilinogen
-
-
-
96.3
100.0
94.1
3.7
-
5.9
-
96.4
3.6
0. 6o
0.71
Corphoporphyrin
29.6
18.2
11.8
70.4
81.8
88.2
-
-
-
21.8
78.2
-
2.05
0.35
Uroporphyrin
66. 7
45.5
41.2
33.3
54.5
58.8
-
-
-
54.5
45.5
-
3.19
0.20
-------
TABLE IX
Percentage Distribution of Laboratory Results: Below, Within and Ahov» Normal Rnnfie
for Kntlie Study Population
I.ab!" "
NeenIive
Trace
+
4-f
Meg.
T rare
+
-l-f
CI. i
SIR.
Trr.t
PBIS
(n-36)
F.
(n-36)
S
(n-36)
0
(n=30)
PBB
(n=36)
E S
(n=36) (n~36)
0
(n=30)
PUB
(..=36)
E
(n=36)
S
(n-36)
0
(n=30)
PBB
(n~36)
F.
(n=36)
S
(n-36)
0
(n~ 30)
Totals (138 persons)
Square
l.e v.11
Urine 0i[.mick
of Spot
Urine
s
N i I r n t e
86.1
100.0
91. 7
fij. J
-
-
-
13.9
-
8.3
16.7
-
-
-
-
90.6
-
9.4
-
6.48
0.09
pi.
-
-
-
-
-
-
-
100.0
100.0
100.0
100.0
-
" -
-
-
-
-
100.O
-
-
Protein
JO. 6
10. 6
JO. 6
46.7
66. 7
63.9 66.7
46.7
2.8
5.6
-
6.7
-
-
2.8
-
34.1
61.6
3.6
0.7
8.55
0.4 7
Cliico.se
91. 7
97.2
97.2
96.7
5.6
-
-
2.8
-
2.8
-
-
2.8
-
3.3
95.7
1.4
1 .4
1.4
9.79
0.36
Ke roues
88.'>
RR.9
97.?
83.¦*
-
-
3.3
5.6
8.3
-
13.3
5.6
2.8
2.8
-
89.9
0.7
6.5
2 .
10.4 3
0. .1
B loud
94. 4
88.9
91 .7
83.3
2.8
8.3 5.6
-
2.8
2.8
-
10.0
-
-
2.8
6.7
89.9
4.3
3.6
2. 2
12.41
0. 19
Kctj to AMtcv ('T (icmiS •' PUB, Polyhromimted P.iphenyl Workers; B, K1 eel r i c ions; S, Steel Workers; 0, Others Including 2 "unmatched11 electricians; No Cases.
-------
TABLE X
Correlation Between Laboratory Variililes and Lengt n nf Emp I oyracn t of 36 PBB Workers (Male)
Test r Tes t r
Blood Chemistry Endoc_r ine
Sodium
-0.201
Tes tosterone
0.377
Potassium
-0.195
FSH
0.336
Chloride
-0.021
T,
0.096
C°2
4
0. 103
FT I
0.047
Blood urea nitrogen
-0.199
LH
0.389
Creatinine
-0.074
TSH
0.042
Glucose
-0.089
RUR
0.127
Calcium
-0.158
Phosphorus
-0.129
Total Protein
-0.319
Hematology
Uric Acid
-0.197
Acid phosphatase
0.008
Hemoglobin
0. 138
Cholesterol
0.040
Hematoc rit
0. 120
Bilirubin
-0.177
RBC
0.174
SGPT
0. 125
MCVC
0. 168
yGT
-0.j085
MCH
0. 142
LDH
0.21
MCHC
0.081
A1bumi n
-0.127
Leucocytes
0.123
Globu1 in
-0.221
Neutrophils
0.058
A/G ratio
0. 137
Band
0.110
SG0T
-0.021
Eos inophil
0.272
Alkaline phosphase
-0.090
Basophil
0.491
Triglycerides
-0.290
Monocytes
0.011
Lymphocytes 0.079
I r[ - 0.275, 0.325, 0.381 are the 0.10, 0.05, and 0.02 percentage points (respectively, with 35 degrees
of freedom) of the correlation coefficient under the null hypothesis of no correlation, i.e., £=0.
-------
TABLE XI
Medical, Drinking and Smoking History of Persons with Elevated Liver Function Tests]
Occupational^—
Group
Abnormal Liver
Function Tests
Other data from
laboratory, medical
Alcohol Use
Smoking
-^"^Case
Age
Sex
YGT SGPT SGOT
Bili- LDH Alk
rubin P
history or physical
examination
Use of
Drugs
per
B
week2
W Md
History
PPD3
Yrs.
PBB Workers
1
51
M
X
cirrhosis of liver,
+BP, kidney stones,
present +TSH
Heavy
in past
lh
38
2
26
M
X
bromide fumes in eyes
24
occ.
1
7
3
26
M
X
2
-
1
2
4
58
M
X
malaria, MI, shingles
aspirin,
Tolinase
1
- -
pipe,
6-8
5
31
M
X X
X
ex-heroin addict
Methadone
14
_
1
7
Electricians
6
43
M
X
48
10
3
22
7
27
M
X
hip fract., pneumonia
2
1-2
lh
8
8
48
M
X X
glomerulonephitis,
pneumoni a
-
6 2
2
25
9
56
M
XXX
X
lymphoma
radiation
Compazine
4
3
1
3
Steel Workers
10
23
M
X
100
20
1
5
11
61
M
X
hemorrhoids, hernia
-
-
-
-
124
30
M
XXX
X
serum hepatitis,
bronchitis,asthma
Bronkade
42
-
1
13
13>4
27
F
X X
X X
chronic hepatitis
(serum hepatitis)
35
35 65
2
8
yGT >75 y/1; SGPT >42 y/1; SGOT >595 y/1; Bilirubin >1.275 mg/dl; LDH >375 y/1; Alkaline Phosphatase >45.5 y/1.
2B, cans of beer; W, glasses of wine; Md, mixed drinks.
^PPD, packs of cigarettes per day.
^Cases 12 and 13 are husband and wife.
-------
TABLE XII A
Total PBB's Serum Levels by Occupational Group
Results
PBB
Workers
Electricians
Steel
Workers
Others
Tot a 1
No
%
No.
%
No.
%
No.
%
No.
%
Not detectable
13
33.3
20
47.6
13
50.0
16
53.3
62
45.3
Not confirmed*
9
23. 1
14
33.3
7
26.9
11
36.7
41
29.9
Trace"""
3
7.7
3
7.1
2
7.7
0
0.0
8
5.8
Detectable level
14
1 35.9
52
11.9
43
15.4
34
10.0
26
19 .0
Not available
0
_
0
—
0
—
1
_
1
—
Total
39 100.0
42 99.9
26
100.0
31
100.0 138 100.0
Not confirmed = impossible to confirm any detectable level. For most practical purposes,
equivalent to "not detectable".
** Trace = below measurable level
Not included in the total
1
Serum levels of PBB workers: .5, 2.4, 12.0, 15.0, 20.0, 20.5, 29.0, 40.4, 50.0, 58.0, 94.0,
132.3, 318.0, 1340.0 (ng/mL)
Serum levels of electricians: .4, 1.9, 2.0, 4.0, 6.0 (ng/mL)
Serum levels of steel workers: .3, 1.5, 2.2, 7.8 (ng/mL)
Serum levels of others: 2.2, 5.5, 13.0 (ng/mL)
REMARK; A significantly higher number of PBB workers had a detectable level of PBBs as compared with
other workers in the study (35.9 percent as compared to 12.2 percent, chi square = 8.67,
p<0.005). Among workers with detectable PBB levels, the PBB workers had significantly higher
serum levels (p<0.001) as compared to other workers (Z = 3.61, Mann-Whitney U test).
-------
TABLE XII B (continue^)
Total PBBO's Serum Levels by Occupational Group
Results PBB workers Electricians Steel Workers Others ToLal
No. % No. % No. % No. % No. %
Not detectable 14 35.9 8 20.0 6 24.0 10 37.0 38 29.0
Not confirmed" 6 15.4 10 25.0 8 32.0 7 25.9 31 23.7
Trace** 7 17.9 6 15.0 6 24.0 4 14.8 23 17.6
Detectable level 121 30.8 162 40.0 53 20.0 6A 22.2 39 29.8
Not available 0- 2- 1- 4-7-
Total 39 100.0 42 100.0 26 100.0 31 99.9 138 100.1
" Not confirmed = impossible to confirm any detectable level. For most practical purposes
equivalent to "not detectable".
Trace = below measurable level
Not included in the total
^Serum levels of PBB workers: .2, .4, .5, .6, .8, .9, 1.3, 2.3, 2.4, 2.7, 5.2, 5.2 (ng/mL)
^Serum levels of electricians: .4, .5., .7, .8, .8, .8, .9., 1.0, 1.0, 1.3, 1.6, 1.6,
2.0, 3.1, 3.7, 4.0 (ng/mL)
3
Serum levels of steel workers: .3, .6, .6, 1.8, 2.1, 2.5 (ng/mL)
4
Serum levels of others: .7, .8, .9, 1.9, 2.8 (ng/mT.)
-------
TABLE XII C (contined)
Total PCB's Serum Levels by Occupational Group
Results
PBB
No
workers
%
Electricians
No. %
Steel
No.
Workers
%
Others
No. %
Total
No. %
Not detectable
2
5.1
1
2.5
0
0.0
0
0.0
3
2.3
Not confirmed*
3
7.7
2
5.0
1
4.0
4
14.8
10
7.6
Detectable level
34
1 87.2
372
92.5
243
96.0
234
85.2
118
90.1
Not available
0
-
2
-
1
-
4
-
7
-
Total
39
100.0
42
100.0
26
100.0
31
100.0
138
100.0
* Not confirmed = impossible to confirm any detectable level. For most practical purposes
equivalent to "not detectable"
Not included in the total
The 4 highest levels of PBB workers: 74.7, 71.2, 31,7, 30.4 (ng/mL)
2
The 4 highest levels of electricians: 5,321.8, 349.0, 77.3, 34.0 (ng/mL)
3
The 3 highest levels of steel workers: 49.6, 41.7, 40.4 (ng/mL)
L
The 3 highest levels of others: 33,000, 194.0, 31.8 (ng/mL)
-------
TABT.E XIII
Detectable" Serum Levels of PBBs, PBBOs and PCBs (nf^/mL)
(138 Samples Analyzed)
Polybrominated Biphenyls (PBBs)
No. Minimum Maximum
C12H9Br
C12H8Br2
C12H7Br3
C12H6Br4
C!2H5Br5
C12H4Br6
Cl2H3Br7
C12H2Br8
C12HBr9
C12BriO
14
1
1
0
2
2
7
9
1
0
0.3
6.9
0.9
1.6
0.4
9.0
20.0
500.0
5.5
6.9
0.9
13.0
6.0
40.0
800.0
500.0
Polybrominated Biphenyl Oxide (PBBOs)
No. Minimum Maximum
C12H9BrO 36
C12H8Br2° °
C12H7Br3° 8
C12H6Br4° °
C12H5Br50 0
C12H4Br6° °
C12H3Br70 °
C12H2Br8°
C12HBr90
C12Brl0°
0
0
0
0.2
0.8
3.1
.4.7
Polychlorinated Biphenyls (PCBs)
No. Minimum Maximum
0.6 14.0
c12h9ci
C12H8C12
C12H7C13
'12H6C14
C12H5C15
C12U4C16
C12H3C17
C12H2C18
c12HC19
9
10
48
101
103
105
54
1
0
0.3 2.3
0.2 240.0
0.3 7,600.0
1.0 20,000.0
0.6 4,900.0
0.3 200.0
27.0 27.0
Total PBBs 26
0.3 1,340.0
Total PBBOs 39
0. /.
5.2
Total PCBs 115
1.0 33,000.0
^Excludes cases with "trace", "not detectable" or "not confirmed" levels, and cases where sera were "not available"
for testing
rs
-------
67
Table XIV
Selected CharC'Cteris ties and Cause of Death
for Decedent P33 Workers
w o v k e r
Worker 3
verier
worker D
Age
Occupation
Dates of
Employment
Dates of
Death
Ir-mediat e
Cause of
Death
Cont r ibutir.p
Cause of
Death
3m
Chemical
opera to r
11/7/69 to
1/4/76
]. / 4 / 7 6
Acute nyocard-
ial infarction
due tc coronary
sclerosis
von ?eckl ir.g-
hausen's
disease
63
Maintenance
5/20/71 to
4/4/76
4/4/76
ruetas tatic
c roncriogenic
carcinoma
Chemical
operator
2/30/70 to
56
Maintenance
2/26/69 to
12/2/77
5/16/78
electrolyte
imbalance due
to peritcxi:is,
associated with
passive arcites
due to partial
cirrhos is
diabetes
tr.ellitus ,
arterio-
sclerotic
heart disease
12/2/17
metastatic
lun2 cancer
Autopsy
No
•NO
Yes
Because a total cf only four known deaths has occurred, a meaningful comparison
could not be made with expected deaths. However, it should be noted that all
workers died at relatively voun?, apes. Two workers died o? the cancer of the
respiratorv system, a disease often cited as occupational risk in a chemical
dust-billed environment, and two workers had heart disease. Although there is
no wav to determine whether these workers had a thvroid dvsfunction, it is
recognized that hypothyroidism can lead to atheroscleratic heart disease.
------- |