v>EPA
          United States
          Environmental Protection
          Agency
          Health Effects Research
          Laboratory
          Research Triangle Park NC 27711
EPA-600/1-78-037 a
May 1978
          Research and Development
Human Scalp Hair:
An Environmental
Exposure Index for
Trace Elements

I. Fifteen Trace
Elements in
New York, N.Y.
(1971-72)

-------
                RESEARCH REPORTING SERIES

Research reports of the Office of Research and Development. U.S Environmental
Protection Agency, have been grouped into nine series These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields
The nine series are1

      1   Environmental Health  Effects Research
     2.  Environmental Protection Technology
     3   Ecological Research
     4   Environmental Monitoring
     5   Socioeconomic Environmental Studies
     6.  Scientific and Technical Assessment Reports (STAR)
     7   Interagency Energy-Environment Research and Development
     8   "Special" Reports
     9   Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL HEALTH EFFECTS RE-
SEARCH series  This series describes projects and studies relating to the toler-
ances of man for unhealthful  substances or conditions. This work is generally
assessed from a medical  viewpoint, including physiological or psychological
studies  In addition to toxicology and other medical specialities, study areas in-
clude biomedical instrumentation and health research techniques utilizing ani-
mals — but always with intended application to human health measures
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.

-------
                                                    EPA-600/l-78-037a
                                                    May 1978
            HUMAN SCALP HAIR:   AN ENVIRONMENTAL
             EXPOSURE  INDEX  FOR TRACE ELEMENTS

I.   Fifteen Trace Elements in  New York,  N.  Y.  (1971-72)
                            by

                      John P.  Creason
           Statistics and Data Management Office
            Health Effects Research Laboratory

                           and

                     Thomas A. Hinners
                    Joseph E.  Bumgarner
                      Cecil Pinkerton
      Environmental Monitoring and Support Laboratory
           U.S. ENVIRONMENTAL PROTECTION AGENCY
            OFFICE OF RESEARCH AND DEVELOPMENT
            HEALTH EFFECTS RESEARCH LABORATORY
            RESEARCH TRIANGLE PARK, N.C. 27711

-------
                                 DISCLAIMER
     This report  has  been  reviewed by the Health Effects Research Laboratory,
U.S. Environmental Protection Agency, and approved for publication.  Mention
of trade names or commercial products does not constitute endorsement or
recommendation for use.
                                    ii

-------
                                FOREWORD
     The many benefits of our modern, developing, industrial society
are accompanied by certain hazards.  Careful assessment of the relative
risk of existing and new man-made environmental hazards is necessary
for the establishment of sound regulatory policy.  These regulations
serve to enhance the quality of our environment in order to promote the
public health and welfare and the productive capacity of our Nation's
population.

     The Health Effects Research Laboratory, Research Triangle Park,
conducts a coordinated environmental health research program in toxicology,
epidemiology, and clinical studies using human volunteer subjects.
These studies address problems in air pollution, non-ionizing
radiation, environmental carcinogenesis and the toxicology of pesticides
as well as other chemical pollutants.  The Laboratory participates in
the development and revision of air quality criteria documents on
pollutants for which national ambient air quality standards exist or
are proposed, provides the data for registration of new pesticides or
proposed suspension of those already in use, conducts research on
hazardous and toxic materials, and is primarily responsible for providing
the health basis for non-ionizing radiation standards.  Direct support
to the regulatory function of the Agency is provided in the form of
expert testimony and preparation of affidavits as well as expert advice
to the Administrator to assure the adequacy of health care and surveillance
of persons having suffered imminent and substantial  endangerment of
their health.

     These data are provided for those researchers interested in developing
a reliable and easily collected index of environmental exposure to certain
trace elements, and as well, the data shed light on  the influences of
personal  covariates on the trace element content of hair.  These data are
timely with regard to the current concerns regarding low-level environmental
exposure to trace elements and their uptake by exposed populations.
                                   F. G. Hueter, Ph. D.
                                     Acting Director,
                           Health Effects Research Laboratory
                                    iii

-------
                                  ABSTRACT

     Previous studies have revealed that hair trace element concentrations
can reflect exposure in cases of frank poisoning and deficiency.   Corre-
lations have been found also in some populations living in regions where
metallurgic processes are conducted.
     This study reports significant correlations between hair barium,
chromium, lead, mercury, nickel, tin, and vanadium content and exposures
(as measured by analyses for the corresponding elements in dustfall or
housedust) within a single metropolitan area.  Age, sex, hair color, and
smoking habits were included in the statistical evaluation.  Several
metals showed a tendency to increase and decrease together in the hair
specimens in agreement with trends reported for other human tissues.
     It is acknowledged that hair has the capacity to adsorb and to
release trace elements in certain situations.  However, population studies
can compensate for confounding influences by (1) a randomizing effect,
by (2) an averaging effect, and (3) by statistical rejection of unrepre-
sentative data values.  The relationship of hair content to (a) content
in other tissues and to (b) metabolic status are separate and complex
issues that should not be confused with (c) exposure relationships.
                                     IV

-------
                                  CONTENTS

                                                                          • • •
Foreword	1!1
Abstract 	   ™
Figures	Y1.
Tables	V.V.
Acknowledgment 	v111

    1.  Introduction 	    J
    2.  Methods	    J
             Environmental Monitoring	    •}
             Scalp Hair and Covariate Information	    *
             Chemical Analysis 	    |
             Statistical Analysis	    °
    3.  Results	    °
             Study Population Characteristics	    °
             Sample Hair Trace Metal Characteristics  	    8
             Pollution Media Trace Metal  Characteristics  	    9
             Hair Trace Metal Concentrations  in Relation
                to Media  Indices  of Trace Metal Exposure
                and to Personal Covariates	'1
    4.  Summary and Discussion 	   15

References  	

-------
                                  FIGURES
Number                                                                 Page
  1   Trace metal levels by element and community for
       dustfall	20
  2  Trace metal levels by element and community for
       housedust	21
  3  Trace metal levels by element and community for soil  	  23
  4  Geometric mean scalp hair Pb for children by age and
       sex	24
  5  Geometric mean scalp hair Mn for children by age and
       sex	25
  6  Geometric mean scalp hair Ba for children by age and
       sex	26
  7  Geometric mean scalp hair Ba for adults by age and
       sex	27
  8  Geometric mean scalp hair Cu for adults by age and
       sex	28
                                      VI

-------
                                  TABLES


Number                                                                 Page

  1   Sample Preparation and Analytical  Methods 	   29

  2  Number of Participants by Age, Sex, and Area of
       Residence	30

  3  Demographic Characteristics of Adults by Area of
       Residence	31

  4  Hair Color by Age Category and Sex	32

  5  Trace Metal Levels in Human Hair in Children
       (Ages 0-15) and Adults  (Ages +16) (ug/g)	33

  6  A Significance Table for  Hair Element/Element
       Correlations	34

  7   Trace Metals in Dustfall  by Area (mg/m2/mo)  	   35

  8   Trace Metals in Housedust by Area  (yg/g)	36

  9   Trace Metals in Soil  by Area (yg/g)	38

 10   Correlation Coefficients  of Logs of Media Trace
       Metal  Levels	39

 11   Tests of Significance of  the Effect of Selected
       Factors on Soil  Trace Metal  Levels	40

 12   Tests of Significance of  the Effect of Selected
       Factors on Scalp Hair Trace Metal  Levels,  Using
       (1) Dustfall  and (2) Housedust as a Measure of
       Environmental  Exposure	41

 13   Dustfall, Housedust,  and  Scalp Hair Mean Trace Metal
       Levels by Community for Scalp Hair Metals  with a
       Significant Environmental  Exposure Effect  	   42

 14   Geometric Mean Scalp Hair Trace Metal Concentrations
       by Sex for Children and Adults (yg/g)	43
                                     VII

-------
                               ACKNOWLEDGMENT

     The authors gratefully acknowledge Ms. Peggy Stewart, Dr. Anna Yokum
and the personnel of Stewart Laboratories,  Inc., Knoxville, Tennessee,
for their major contributions  and dedication to this study.
                                     vm

-------
                                  SECTION 1
                                 INTRODUCTION

     Human scalp hair has been shown to reflect increased environmental
exposure to metals such as lead, mercury, cadmium and arsenic (1-6).
Hair chromium and zinc concentrations have been reported to reflect de-
ficiency conditions in humans (11,22,38,40-46).  Similar trends for
these and other elements have been found in animal  studies (11,38,43,44
47).  Scalp hair is an almost ideal  tissue for population sampling in
that it is painlessly removed, normally discarded,  easily collected and
easily stored (1,2,38).  The CHESS* program of the  United States
Environmental Protection Agency (7) provided a means for community-wide
sampling of scalp hair.
     The purpose of this study was to evaluate further the utility of scalp
hair as a method of environmental monitoring of humans for trace metals
exposure.  The relationship of scalp hair trace metal levels to important
personal covariates such as hair color, age, sex, and socioeconomic level
was also of concern.  Since for some metals, smoking has been implicated
as an important exposure covariate (8,9), it was included in this study.
A knowledge of the effects of these personal covariates on scalp hair trace
metal levels will permit a more accurate assessment of the quantitative
relationships of hair trace metal levels to environmental exposure.
     Dustfall has been used as an environmental index of trace substance
exposure (19,48).  More intimate indices of trace substance exposure such
as household dust, soil and water from CHESS-participant homes have already
been considered (9,10).
*CHESS stands for Community Health and Environmental Surveillance System.
                                      1

-------
     The specific hypotheses  tested in this study were:
     1.  Significant relationships exist between dustfall, household
         dust  (housedust), and  soil trace metal levels.
     2.  There are significant  variations in selected scalp hair
         trace metal levels due to personal covariates such as age,
         sex, hair color, socioeconomic status and smoking habits.
     3.  Environmental exposure, as measured by one of the above media,
         has a significant effect upon selected scalp hair trace metal
         levels, even after adjusting for any effects due to personal
         covariates.
     One covariate of great interest that could not be evaluated in this
study was race.  There were too few non-white respondents to make such
an investigation possible.  This covariate will be examined in later
studies.
     Other than the above specific hypotheses, it was of interest to
establish the distributional  characteristics of each scalp hair trace
metal, including baseline levels, ranges, and skewness of the distributions.
     Although measurements of nineteen (19) different scalp hair trace
elements were attempted, only the following fifteen (15) will be dealt  with
in this report:  barium  (Ba), boron (B), cadmium (Cd), chromium (Cr),
copper (Cu), iron (Fe),  lead  (Pb), lithium (Li), manganese (Mn),
mercury (Hg), nickel (Ni), selenium (Se), silver (Ag), tin (Sn) and
vanadium (V).  Arsenic (As),  beryllium (Be) and cobolt (Co) were excluded
from consideration because many hair sample values were below detection
limits for the sample sizes available.  Zinc (Zn) was measured in all
three environmental samples collected as well as in scalp hair, but the scalp
hair concentrations were found  to be low in comparison to most normal values

-------
in the literature (11,22).  A report on this metal has therefore been
postponed.  Five metals (Cd, Cu, Pb, Mn and Ni) were measured in all the
media (dustfall, housedust, and soil) and the remaining nine metals were
measured in housedust only.

-------
                                  SECTION 2
                                   METHODS
Environmental Monitoring
     Three CHESS communities were selected in the New York City metropolitan
complex:  the Westchester section of Bronx, the Howard Beach section of
Queens, and the town of Riverhead, Long Island.  These communities were
believed to exhibit a general pollution exposure gradient from low
(Riverhead)  to  intermediate  (Queens) to high  (Bronx); this belief was
based  on previous  observations  of  total suspended particulate  levels and
dustfall levels,  as well  as  some preliminary  examinations of levels of
selected trace  metals  in  dustfall  (10,12).  Atmospheric  studies  including
dustfall collections,  were made at CHESS  air  monitoring  sites.  These
 sites were located such that the families from each of  the  three communities
 of interest were within 2.5 km. of the respective sites.
      Dustfall data were obtained monthly over a period  of 22 months, from
 September, 1970 through June,  1972, at the central  site within each
 community.  Standard procedures of collection were followed (13).  Four
 soil  samples were obtained  from each of  43 residences (14 each from
 Queens and  Riverhead and  15 from  Bronx).  Two samples were collected
 from  the  front yard, and  two from the  back yard.   All samples were  taken
 to  a  depth  of  two inches  with  a stainless  steel  auger.   Ninty-nine
 housedust samples (29  from Queens, 37  from Riverhead,  and  33  from Bronx)
 were obtained  by collecting the contents of vacuum cleaner dust bags  from
  community homes.
  Scalp Hair and Covariate Information
       In March, 1971, letters  giving information about the proposed trace
                                        4

-------
 metals study were sent to all  active CHESS acute respiratory  disease  (ARD)
 families in the three communities.   These letters included  for  each family
 member a return addressed postpaid  envelope,  with the individual member
 identification label  affixed  to the back, and instructions  for  the
 collection of the hair sample.
      In the instructions  it was stressed  that the hair should be taken
 from the next normal  haircut  or trim,  and that as much hair as  possible
 should be collected.   It  was  also stressed that there was no special  need
 to wash or shampoo the hair before  the haircut or trim, because it would be
 washed in the laboratory  before analysis.   Polyethylene bags for hair
 samples were  sent with the letters  and had an identification label on the
 back with each  person's first name,  family name and a  space for the date
 of the haircut.   During a scheduled  ARD survey phone  call,  after allowing
 sufficient  time for receipt of  the  letters, the color  of each member's
 hair and  the  location  of the haircut (barbershop  or  home)  were ascertained.
 This information  was then combined with the CHESS  ARD  background information
 questionnaire obtained at the start of the ARD study  to make a  complete
 covariate information  file on each contributing family member.  Collection
 of  hair samples was terminated  in July 1971.   The  hair samples  were then
 stored  from July  1971  until trace metal analyses were carried out in  the
 Spring  of 1973.
 Chemical Analysis
     Dustfall samples were acid extracted and  the metals determined by
 atomic absorption spectrophotometry  (13).   Housedust samples were sieved
 through a 0.5 mm  screen for 5 minutes at 260 oscillations per minute  on
 a mechanical shaker, and extracted with 6N nitric acid at 50°C  for 30
minutes.  Soil samples were also sieved through a 0.5 mm screen.  Twenty
                                       5

-------
grams of soil was  then  extracted with 40 ml of IN HC1.  Hair specimens
were washed with a detergent  solution, rinsed and dried according to the
procedure of Harrison et al.  (14).   Digestion was achieved by oxygen
combustion for some elements  and by  dry ashing for others, as indicated in
Table 1.  To prevent any losses when volatile elements were to be
analyzed, weighed  portions  of the washed and dried hair were prepared
by the Schoniger Flask  technique.
     Analytical methods for each metal are also shown in Table 1.  Standard
laboratory quality control  procedures were employed.  In addition, recovery
of all 19 elements added to a housedust sample and to a hair sample were
evaluated using additions that were  either twice the detection limit or
twice the endogenous level, whichever was larger.  Recovery rates were
greater than 85 percent in  all cases, and greater than 95 percent in most
of the cases.
Statistical Analysis
     The first step taken in  preparation for statistical analysis of the
hair data was the careful editing of the data for outliers, values so
far removed from the main body of readings as to warrant their removal
from the population for statistical analysis purposes.
     A statistical procedure  was developed for this process, so that all
subjective tendencies toward  removal of any values were eliminated.  In
this procedure the inherent population variability, as measured by the
standard deviation of the logs of the values, was estimated from the
central  section of the  sample.  Limits were then obtained by taking +3
standard deviations from the  mean of the logs of the sample.  Histograms
of the data were carefully  examined  to insure the effectiveness of this

-------
 procedure, and to verify that a large number of seemingly valid obser-
 vations were not being eliminated.  No such problems were ever encountered.
 For the trace metals measured, the percent declared outliers varied from
 0.2 percent for lithium up to 7.6 percent for selenium, with the median
 percent rejected being 1.5 percent.
     Examination of the trace element values revealed a consistent
 tendency toward log-normality of the distribution (i.e. populations skewed
 to the right), so that logs of all values were employed in the subsequent
 statistical analyses.   Standard statistical techniques of correlation and
multiple linear regression were then used to discover the effects and
 interrelationships of  all  of the measured variables.

-------
                                    SECTION  3
                                     RESULTS
 Study Population  Characteristics
      Over 3000  subjects  participated in the CHESS-ARD study (13);  426
 families, comprised of about  1900  total members, responded to the  hair
 study letter in some fashion.  Family members who gave no hair,  who
 gave an insufficient quantity of hair, or who had incomplete covariate
 information, were excluded from the study.   Because responses were
 available from very few nonwhite families,  these persons were also
 excluded.   The resultant population consisted of 498 subjects with
 complete information.   The distribution of  subjects by age,  sex  and
 residence  is shown in  Table 2.  The study population is  seen  to  consist
 of two  distinct age groups.  This  is a  result of the method of contact
 in the  ARD  study,  wherein only families having children  in elementary
 schools  were selected.   This  population division presents no  difficulty,
 however, since  subsequent analyses  are  made on children  and adult
 populations  separately.   The  distribution of  the final group  of  participants
 with respect to  smoking  patterns and education for  head  of households
 (Table 3) was similar to  the  original ARD population  (13).
     Reported hair color  by sex for children  (15 years of age and less)
 and adults (over 15 years of  age) showed significant differences in hair
 color distribution between  male adults  and male children, but not between
 female adults and  female children (Table 4).   The adult males showed a
higher percentage of black  and grey  hair than male children.
Sample Hair Trace Metal  Characteristics
     The 15 trace metals in this study all  have typical log-normal  type
                                      8

-------
of distribution usually displayed by trace metals  in  hair  (15,16).
Analytical hair values obtained generally agree  with  published  values
(3,11,15,16).
     The interrelationships of scalp hair trace  metal  levels  were
examined separately for children (15 years of age  or  less) and  adults
(over 15 years of age), since children are in a  rapidly developing
stage of growth in contrast to adults who have reached a leveling
off in growth and development.  Some adults also are  likely to  have
been exposed to some work-related sources of trace metals, and  hence
may display a good degree more variability from person to person in
scalp hair trace metal burden than children.
     It was found that adults have higher mean scalp  hair levels than
children in 10 out of the 15 metals (Table 5).
     Pairwise correlation coefficients were computed  separately for the
adults and children groupings using the logs of scalp hair trace metals.
There were 68 significant correlations for adults and 85 for the children
out of the 120 correlation coefficients  (Table 6).  Selenium (Se) was
the only element to consistently show significant negative correlations.
In both adults and children Cd, Cu and Pb were highly  inter-correlated.
     The relation of  hair trace metal levels to media  indices  of trace
metal exposure and to personal covariates are considered  in  a  later
section of this  report.
Pollution Media  Trace Metal Characteristics
     Arithmetic  means of dustfall  and geometric means  of  soil  and
housedust trace  metal  concentrations  by  community are presented in
Tables  7, 8  and  9.

-------
     A  previous  study  of pollution  data  from  these same sites within
these same  communities  established  the fact that  there were indeed
statistically  significant differences in trace metal levels between
communities  in dustfall, soil,  and  housedust  for  Cr, Cu, Mn, Ni, Pb
and Cd, except that  chromium was  not analytically determined in the
soil, and Cd was  found  to display no difference in the housedust levels
across  communities  (10).  Bar graphs of  the means for dustfall and
geometric means  for  housedust and soil by community are shown in Figures
1-3.  These figures  demonstrate the concentration gradients as well as
the similarities  and differences  in trace metal patterns across these
three media.  Cd, Cr and Ni  generally have much lower concentrations than
Cu, Pb, Mn and Zn.   There is apparently  a higher  percent of Mn in the
soil than in housedust  or dustfall, while dustfall seems to have relatively
less Ni and Cu than  the other two environmental media.  For most of the trace
metals the three  media  indices  of exposure are quite similar, reflecting
increased exposure  in moving from Riverhead to Queens to Bronx.  In order to
more precisely assess the interrelationships  of soil, housedust and dustfall,
correlations of  the  respective  trace metals across these media were
computed (Table  10).  Logs of the concentrations  were used in this
computation in order to normalize the data and make significance tests
valid.  One should  keep in mind that all  correlations with dustfall
metals involve the  pairing of housedust  and soil  metals within a community
to a single dustfall value.   However, this correlation produces the same
results as one would obtain  by  a  simple  linear regression of the media
metal on the dustfall metal.  The housedust-soil  correlation was
obtained by first averaging  the four soil values  to obtain a single
value per home, and  using all houses for which both housedust and solid
                                      10

-------
results were available In  order to  obtain  the  correlation coefficients.
All of the correlation coefficients were found to  be  highly  significant
indicating a strong relationship between these measures  of environmental
trace metal levels for the six metals available for analysis.
     The use of soil as an environmental index of  trace  metal  burdens
has several problems that  have been mentioned  in earlier reports (19,20,
21).  The problems that must be considered are background levels in the
soil, possible gradients in soil concentrations due to roadway traffic,
and contamination from older homes  on which a  great deal of  lead-based
paint had been used.  From an analysis of the  logs of soil  trace metal
levels (Table 11), the strong relationship of  dustfall levels to soil
levels  is seen, as well as the great variability  between homes within the
areas, and between the front and back yard measurements.
Hair Trace Metal Concentrations in  Relation to Media Indices of Trace
Metal Exposure and to Personal Covariates
     One must take into account as  many influences on scalp hair trace
metal levels as possible in order to assess adequately  its utility
as an epidemiological personal exposure index.  The first step  in this
process was the separate analysis of children  (5!5 years old) and adults
(>J6 years old).  Within each of these  groups, the following possible
influences were assessed:  age, sex, hair color, location of haircut
(home or  barbershop),   socioeconomic level  (as measured by education  of head
of household) and  smoking  patterns  (in  adults  only).  The relation
of environmental  exposure  to  hair  trace metal  levels, after  adjusting
for  the effects due  to  these  covariates, was  then  tested using  a  linear
multiple  regression  model.  The logs of both  the  scalp  hair  trace metal
                                      11

-------
levels and the environmental exposure trace metal  levels were used  in
the analysis to make the scalp hair values more closely fit a normal
distribution and to help insure the fit of a linear model to the  data.
     The results of the statistical tests when monthly average trace
metal dustfall rates were used are given in Table 12.  Dustfall  trace
metals had a significant effect on hair levels for Pb, Cr, and Ni in
children but only on hair levels of Pb in adults.  All of these scalp
hair metals show a marked increase in concentration with the media
gradient (Table 13).
     Housedust was used in  place of dustfall as an environmental  index
in the above models by computing the geometric mean of the housedust
trace metal readings obtained  from contributing households within each
community.  The results of  this substitution were essentially identical
to those with dustfall in the  model, except that housedust chromium was
not  indicated as having a significant effect on scalp hair chromium in
children (Table 12).
     By selecting  the  family members of  households contributing  housedust
and  computing the  correlation  of these scalp hair trace  metals and the
corresponding housedust trace  metals, a  more direct  comparison of
environmental exposure and  scalp hair levels of  trace metals  is  possible,
although other  covariates must of  necessity be ignored.   Significant
correlations  found were  for Pb (r=0.27)  and Ni (r=0.32)  in  children
and  for Cu (r=0.31)  in adults.  It is apparent that  the housedust  and
dustfall  trace  metal  values compared above can be  used  interchangeably
as  an  exposure  index.
                                      12

-------
     There are nine metals  in  this  study  for which the only pollution
index available is housedust:   Ba,  B,  Fe, L, Hg, Se, Ag, Sn and V.
The linear model  analysis for  these metals using housedust as an environ-
mental index is shown in Table 12.
     Selenium was found to  show no  differences  in housedust trace metals
across communities, so that the effect of selenium in  housedust on  scalp
hair selenium levels could  not be tested  here.  Ba,  Pb, Hg and V were
found to have a significant relationship  between housedust and scalp hair
levels for both adults and  children, while Sn was significant for children
only.
     Significant trends of scalp hair  Pb, Mn and Ba  with age were found
in children, while Cu and Ba in scalp  hair were related to age in adults.
In children, scalp hair Pb decreased with age while  Mn and Ba showed an
increase with age.  In adults, both Ba and Cu decreased with age.   A
separate linear model including an  age-sex interaction term was  also
examined in the belief that the concentration change with  age might be
different between sexes.  This proved  to be true  for adult scalp hair  Cu
and Ba as well as for Ba, Pb and Mn in hair of  children  (Figures 4-8).
     When examined separately by sex,  the female  adults were  found  to
have rapidly decreasing Cu scalp hair levels, while male  levels  were
fairly steady over age.  These trends in Cu levels are in close  agreement
with the results of Schroeder et al.   (15).  The increased hair Ba and Mn
values in females  (but not males)  at ages 12-13 is intriguing in reference
to puberty.  Mn does  have a role in reproduction (39), but Ba has
usually been considered nonessential  (11).  Ba is chemically related  to
                                     13

-------
calcium  (Ca), and  the  higher  concentration of Ba in female hair than in
male hair may be associated with a similar differential observed for Ca
(16,17,36,38).
     For adults, significant  differences in scalp hair trace metals between
sexes were noted for all metals except Cr, B, Fe, Se, Ag and V.  For
children only Cr,  B, Fe and Se were not significantly different by sex.
Females  levels were always higher than male concentrations wherever
significant differences were  found in children.  In adults, females were
higher than males  except for  cadmium and lead.
     Fe  levels in  scalp hair  were found to be significantly related to
socioeconomic level as measured by education for both adults and children
with the scalp hair levels following a negative gradient with increasing
education level.   That is, subjects with higher levels of education of
the head of household had lower levels of Fe in their scalp hair.  Fe
in adult scalp hair was the only metal showing a significant relationship
to hair color, with brown and black hair colors having less Fe, while
gray hair had more.
     Hair samples  collected at barbershops as opposed to home collection
were significantly higher for adults in Cd, Pb, Ba, B and V while in
children Ni, Ba and V were higher.  Selenium in children's hair collected
in barbershops were significantly lower than found in hair from children
who had home haircuts.
     Smokers were  found to have significantly higher hair Se and Fe.
Hair Pb was higher for smokers than for nonsmokers, but was not quite
statistically significant (p=0.06).
                                      14

-------
                                  SECTION 4
                           SUMMARY AND DISCUSSION

     While other studies have Illustrated metal  changes  In  hair when
exposure differences have been dramatic, this report indicates  that hair
metal content can reflect exposure trends within a single metropolitan
area.  Several personal  covariates were found to Influence  scalp trace
metal levels strongly and must be taken into consideration  if scalp hair is
to be used as an environmental index.
     In this study, the environmental  exposure gradients of Pb, Ba, Hg and V
(as measured by dustfall or housedust) were significantly associated with
scalp hair trace metal levels in both  adults and children,  while the
environmental exposure gradients of Cr, Ni and Sn were reflected in
children's hair only.  The seven other metals investigated  (Cd, Cu, Mn, B,
Fe, Li and Ag) displayed no such significant associations.   Housedust and
dustfall trace metal values proved to  be usable interchangeably as an
exposure index for the metals that were measured in both media.
     Sex was found to be the most important covariate in the study, being
significantly associated with 11 of the 15 trace metals examined in children
and 9 of the 15 trace metals in adults.  Female scalp hair  values were
higher than male values in all cases where differences were significant
except for Cd and Pb in adults.  The reversal of the general sex trend for
Cd and Pb could be a reflection of work-related exposure of males to these
nonessential metals, although this hypothesis was not examined in this
study.  Adult males reportedly have more kidney Cd than adult females
(69,70) but Pb did not differ by sex in 33 tissues (70).  However, blood
Pb is reported (9) to be lower in women than in men in similar environments.
                                     15

-------
Other investigators have also found sex differences as observed in this
study, with  females most often  higher than males  (3,11,15-18,25,36).   The
tendency  for female hair to  be  higher than male hair in several metals
may  be related  to  a higher inorganic content  for  female hair on the
average  (49).  Additional  studies may also  investigate whether the sex
difference might be  explained by other  factors such as hair length or
 shampoo  frequency.
      Scalp hair Pb values decrease rapidly with age in children.   This
 decrease is  probably a result of a proclivity to pica as  well  as  a higher
 respiratory  rate  in the very young.  Mn and Ba begin increasing in
 children's  scalp  hair around age 8; Ba displays  peak values in women at age
 11-12 and 31-35 while Mn  does  not change with age  in adults.   The causes
 of  these age trends  in Ba and  Mn are not clearly understood.   Cu was found
 to  rapidly  decrease  with  age in adult  females.   Other reports have shown
 a decreasing trend  in  hair copper with age  (15,25,50).
       Hair color and  education of head  of household both  reflected a
  relationship only to scalp hair Fe.   The hair color  differences  in  Fe  were
  significant only for adults.  Over 90 percent of the hair in  this study
  was either  blond or brown, so other hair colors may not have been adequately
  sampled for the  determination of any differences  in trace metal  content.
  Other investigators have reported differences with hair color (11,15,39).
       The increased  concentrations of  Cd, Pb, Ni,  Ba, B and V found for hair
  samples collected  in  a  barbershop or  beauty shop  as compared to  collection  in
  the home may  represent  contamination.   The  concentration of  Se,  however,  was
  found  to be lower  in  hair collected  in barbershops  or  beauty shops  compared
   to the amounts found in children's hair from home haircuts.
                                         16

-------
     Smokers were found to have higher Pb, Fe and Se scalp  hair  levels  than
nonsmokers in this study.   Although these metals have been  reported  to  be
in cigarettes, the authors are unaware of any published studies  in which
the relationship between hair levels and smoking have been  examined,
although blood Pb values have been found to be higher in smokers than in
nonsmokers (9).
     The hair element intercorrelations (Table 6) show agreement with
several element intercorrelations reported for other human  tissues.
Schroeder et al. (52) reported that Cr was significantly correlated  with
Ba, Mn and Ni in 14 to 18 of the 29 human tissues examined.  Ba  and  Mn
correlated in 15 of 28 tissues in another report (53).  In  our data, hair
Ba, Cr, Mn and Ni were significantly intercorrelated in both children and
adults.  As found for a majority of human tissues (54), hair showed  a
significant correlation between Cu and Fe and (in adults) between Cr and
Mn.  Cr and Pb did not correlate significantly in hair (Table 6) nor
in 20 out of 29 other human tissues (52).  Although exceptions to these
trends can be found between the data, hair does not appear to be unique
among body tissues in its compositional variations.
     The highly significant correlations in hair between Cd, Cu  and  Pb
have been found by other investigators (16-18).  As in another study (17),
only hair Se, a nonmetal, showed a negative correlation with certain hair
metals.  The Fe - V correlation in hair is interesting in view of evidence
that V, like Fe, is bound to transferrin in blood (65) and that hemoglobin
and tissue respiration are decreased in Vanadium toxicity  (72).
     Hair growth in humans is a complex cyclic process involving non-growing
and growing phases of variable duration for each fiber plus differences in

                                     17

-------
growth rate depending on age and sex.  In addition scalp hair  grows faster
and is retained longer in summer than in winter (51).
     Metal concentrations are known to differ in hair  from the same
individual and even at different locations along single fibers or fiber
bundles (4,22,23,30-38,40-47,50).  These hair trace metal variations  have
been considered by investigators as evidence for
     (a) external contamination
     (b) body intake variations
     (c) seasonal trends
     The relatively large number (7.6%) of scalp hair Se values rejected
as outliers may be related  to the use of special shampoos containing  Se
(16).  Exogenous  Se on hair was shown by Bate  (23) to be resistant to
removal by a variety of washing procedures.
     Complexity in interpretation of hair trace metal data is increased by
evidence  indicating that  physiological  uptake  by hair of trace elements is
not a  simple growth phenomenon  (36,44).  For this  study, hair was sampled
without restriction as to location  on the scalp or distance from the scalp.
Consequently, effects observed  in the covariate data could possibly be
related to sampling technique such  as differences  in distance from the scalp
between male and  female  samples.  However,  area differences would not be
affected  by these sampling  techniques since variations  in  the sampling
were  randomized  across  all  areas.
      Laboratory  washing  of hair before  metal  analyses  has  been a point of
contention  (2,3,14,22-32,34-38).   A recent  report (36)  indicates that the
binding  of  certain metals to hair is not as strong as  has  been previously
assumed  by  many investigators.   After comparison  testing of  five techniques,

                                       18

-------
the washing procedure described  by  Harrison  et  al.  (14) was selected for
use in this study.   The detergent used  in  this  procedure was demonstrated
to be similar in effect to commercial shampoos, which  agrees with  the
findings of another investigator for non-ionic  detergents  (24).   In an
earlier EPA hair study (1), a metal chelating agent (EDTA) was  included
in the hair washing procedure, but  use  of  this  reagent was not  recommended
in a follow-up report (2).  Other investigators have subsequently concurred
with the opinion that EDTA should not be used in washing  hair  before
analyses for metals (33,34).
     The relationship of hair content to tissue content  and  metabolic
status are separate and complex issues  that should not be confused with
exposure relationships.
     Some evidence indicates that hair trace element content can reflect
whole body content (38,55,58,67) or content in specific tissues (3,38,40,
43-45,50,57,58,64,68).  When hair content does not reflect other tissue
values  (15,38,66,71), hair can reflect the metabolic or health status (38,
43,47,64,68) while the blood and other tissue  values do not (9,38,47,56,59-64,
66).
     The absence of a demonstratable relationship  between hair content
and media values for some  elements  in this  study is not definitive, but
may simply  indicate that:
      (a)  the exposure difference  was not sufficient  for a correlation
          effect on  hair  trace metal levels to be  observed or
      (b)   that  the media  indices employed were not representative
          of the overall  metal  exposure for the population  sampled.
      Future hair trace element  content  reports from other geographic areas
will  help  to clarify the  utility of scalp hair as  a community  exposure  monitor.
                                     19

-------
ro
o
                                     SCALE TO LEFT FOR Cd,  Cr,  AND  Ni;

                                SCALE TO RIGHT FOR Cu, Pb,  Mn,  AND  Zn mg/m2/month
UI

f
4
r
S
z 0.5
«£
0
•^
o



0.1
0

—
—
—
—
—
—



—

rr










Ni












Cu
MHM;












Pb












Mn










Zn
MMM













Cd
RIVERHEAO






Ni



Cf









Cu

Pb
MMEM






Zn





Mi
^••^












QUEENS
^L_
Ui




Cr




Cd






III
^MEBI




















Zn —
Pb
•MM

















Mn














—



——-
—


20.0


f*
e
10.0*
|
f+
£
d

2.0
n
BRONX "
                       FIGURE  1.  TRACE METAL LEVELS BY ELEMENT AND COMMUNITY FOR DUSTFALL

-------
                          SCALE TO  LEFT FOR Cd, Cr, AND Ni;
                       SCALE TO RIGHT FOR Cu, Pb, Mn AND  Zn  yg/g

—
—
«^^M


Cr
__

_ Cd
^^^




t







Pb
•MMM





Zn




Mn
MMB>
••^•i




Cr



Cd





Ni
ill
IM^M

Zn


Pb




Pll
Ml
^^^











a
Zn
^•M









Pb


Cr



Cd










Cu














Mn









—
—
—
~
111
•


	
100
 50
 10
                                                                                     1000
                                                                                     500
                                                                                    100
             RIVERHEAD
QUEENS
BRONX
         FIGURE 2.  TRACE METAL  LEVELS BY ELEMENT AND COMMUNITY FOR HOUSEDUST

-------
                                SCALE TO LEFT FOR B,  Li,  Hg, Ag AND Sn
                         SCALE TO RIGHT FOR Ba,  Fe  X  TO-2,  Se X TO3, AND V yg/g
              30
ro
ro
          OL

          C
          Ul
          o


          "S,   20
01
              10
                             SnBape
                      Li
                         Hg.
                                    Se
                                                     Ba
                                   Li
                                    Hg   Sn
                                                        Fe
                          RIVERHF.AD
                                         QUEENS
                                                                   Li
                                                                           Ba
                                                               Sn
                                                           Hg
                                                                       Aq
                                                                              Fe
                                                                      Se
                                                                              300
                                                                               200
                                                                                         100
                     •2*




                     o

                     <



                     X .
                     0>
                                                                                   I
                                                                                   o
                                                                                   X

                                                                                   
-------
ro
GO
         20
         10
                                   SCALE  TO LEFT FOR Cd  AND Ni;

                               SCALE TO RIGHT FOR Cu,  Pb,  Mn AND Zn Mg/g
                                                       Pb
                                                 Ni
                                                    Cu
                                                              Zn
                        NI
                                                                     Cd
                                                                   n
                                                                               Cu
                       RIVERHEAO
QUEENS
                                                                                  Pb
                                                                                              1300
                                                                                         Zn
                                           200
                                                                                               100
                    £
BRONX
                                           25


                                           0
              FIGURE 3.   TRACE METAL  LEVELS BY ELEMENT AND COMMUNITY FOR SOIL

-------
ro
                   0-1
2-3
4-5
      6-7          8-9




CHILDREN'S AGE IN YEARS
10-11
12-13
14-15
                   FIGURE  4.   GEOMETRIC  MEAN SCALP HAIR Pb  FOR CHILDREN  BY AGE AND SEX

-------
         2.0
ro
CJl
         1.6
      I  L2-
          0.8
          0.4
                  0-1
2-3
4-5
6-7
8-9
10-11
12-13
14-15
                                                     CHILDREN'S AGE IN YEARS
                    FIGURE  5.   GEOMETRIC  MEAN  SCALP  HAIR Mn FOR  CHILDREN BY AGE AND SEX

-------
ro
                 0-1
4-5           fi-7          8-9
     CHILDREN'S AGE IN YEARS
10-11
12-13
14-15
                    FIGURE 6.  GEOMETRIC  MEAN SCALP HAIR Ba FOR  CHILDREN  BY AGE  AND SEX

-------
ro
                                          26-30          31-35         3640         41-45         46-60
                                                ADULT'S AGE IN YEARS
            •ONLY 1 MALE AND 1 FEMALE IN THIS CATEGORY,
             NO ACCEPTABLE MEAN VALUE IS AVAILABLE.

                     FIGURE 7.  GEOMETRIC  MEAN SCALP HAIR Ba  FOR ADULTS BY AGE AND SEX

-------
ro
oo
   55


  49.5


  44jO


  38.5


  33.0
ge

-27.5

"
  22.0


  16.5


  11.0


   5.5
                     \
                     16-20
                       21-25*
26-30
             •ONLY 1 HALE AND 1 FEMALE HERE IN THIS CATEGORY,
               SO NO ACCEPTABLE MEAN VALUE IS AVAILABLE.
  31-35         36-40
ADULT AGE IN YEARS
41-45
4640
                                                                                                               60*
                    FIGURE 8.   GEOMETRIC  MEAN SCALP HAIR Cu FOR ADULTS BY  AGE  AND SEX

-------
     TABLE  1.  SAMPLE PREPARATION AND ANALYTICAL METHODS
   Metals              Preparation                Analysis
Cd, Pb              Oxygen combustion          AA aspiration
Cu, Mn, Zn, Fe      Oxygen combustion          AA aspiration
As, Hg, Se          Oxygen combustion          AA on  vapor
Li                  Oxygen combustion          Flame  photometry
Ag                  Oxygen combustion          ES
Ba, Be, B, Cr       Dry ashing                 ES
Co, Ni, V, Sn       Dry ashing                 ES

Remarks:
1.  Managanese in hair was evaluated from ES data when  detection from AA
    was found to  be inadequate.
2.  AA =  atomic absorption
    ES =  emission spectroscopy
                                29

-------
TABLE 2.  NUMBER OF PARTICIPANTS BY AGE, SEX, AND AREA OF RESIDENCE


0-5
6-10
11-15
16-20
21-25
26-30
31-35
36-40
41-45
46-50
51 +
Total

Riverhead
Mai
5
41
31
7
0
2
7
12
8
6
1
120
es Females
15
25
9
3
1
9
9
10
7
1
2
91
211
Queens
Males
16
27
17
2
0
7
10
10
7
0
2
98
Females
6
20
4
1
2
8
9
8
6
2
3
69
167
Bronx
Males
8
26
8
1
1
0
2
6
7
1
3
63
Females
7
17
5
1
0
5
10
6
6
0
0
57
120
                                30

-------
TABLE 3.  DEMOGRAPHIC CHARACTERISTICS OF ADULTS BY AREA OF RESIDENCE

Smoking Patterns (%)
Never
Ex
Current
River head

37.6
16.5
45.9
Queens

44.2
24.7
31.2
Bronx

36.7
30.6
32.7
Education of Head of Household (%)
     < High School      15.3              10.4            26.5
       High School      48.2              50.6            44.9
     > High School      36.5              39.0            28.6
                                31

-------
                  TABLE 4.  HAIR COLOR BY AGE CATEGORY AND SEX
Adults

Brown
Blond
Red
Black
Grey & White
Unknown
Male
67
10
2
15
7
1
Female
76
17
2
8
4
2
Overall
143
27
4
23
11
3
Children
Male
122
43
6
6
0
2
Female
76
21
3
6
0
2
Overall
198
64
9
12
0
4
Total
102
109
211
179
108
287
                                       32

-------
                  TABLE 5.  TRACE METAL LEVELS IN HUMAN HAIR
                            IN CHILDREN* AND ADULTS*

Children
Barium
Boron
Cadmium
Chromium
Copper
Iron
Lead
Lithium
Manganese
Mercury
Nickel
Selenium
Silver
Tin
Vanadium
Adults
Barium
Boron
Cadmium
Chromium
Copper
Iron
Lead
Lithium
Manganese
Mercury
Nickel
Selenium
Silver
Tin
Vanadium
No. of
obs.
267
265
281
261
279
282
284
277
267
280
265
260
266
265
267

197
197
201
192
204
202
207
206
197
203
194
188
198
191
193

geo. mean
0.762
0.881
0.88
0.56
12.11
20.83
13.47
0.044
0.56
0.672
0.51
0.320
0.205
0.561
0.250

1.41
0.981
0.76
0.62
18.25
22.30
12.21
0.056
0.95
0.774
0.74
0.303
0.165
0.785
0.182

min.
0.054
0.030
0.14
0.076
1.01
2.70
2.12
0.009
0.05
0.048
0.036
0.025
0.007
0.034
0.010

0.121
0.037
0.08
0.06
2.22
3.60
1.96
0.009
0.07
0.050
0.045
0.025
0.007
0.048
0.009

max.
9.29
22.00
6.90
4.80
144.0
152.00
100.0
0.300
12.0
11.30
11.0
1.65
6.20
8.30
2.90

29.00
25.00
8.73
5.30
184.00
177.00
155.00
0.228
11.0
14.00
11.0
1.58
4.30
12.00
2.20

+1 geo.
0.271
.263
0.42
0.26
4.97
10.52
6.07
0.022
0.22
0.236
0.20
0.158
0.066
0.231
0.085

0.419
0.297
0.33
0.26
7.28
10.46
5.13
0.025
0.34
0.276
0.27
0.140
0.049
0.283
0.056

std. dev.
- 2.135
- 2.956
- 1.85
- 1.22
- 29.48
- 41.24
- 29.91
- 0.088
- 1.45
- 1.914
- 1 .30
- 0.645
- 0.637
- 1.361
- 0.738

- 4.77
- 3.23
- 1.74
- 1.46
- 45.75
- 47.53
- 29.07
- 0.083
- 2.67
- 2.17
- 2.07
- 0.653
Off fm f*
.550
- 2.17
- 0.588
* Children defined as ages 0 through 15, adults as ages greater than 15.
  measurements in
All
                                       33

-------
                           , TABLE 6

 A SIGNIFICANCE TABLE FOR HAIR ELEMENT/ELEMENT CORRELATIONS
     Pb
Cd
Cu
Hg
Li
Se
Fe
Ba
B
Cr
Ni
Ag
V
Sn
Mn
Zn
Pb
Cd
Cu
 Hg
 Li
Se
 Fe
Ba
 B

Cr
                                       •f
 Ni
Ag

 V
Sn
Mn
Zn
+ indicates significant positive correlation and
- indicates significant negative correlation, where significant
  means p<0.05.
                             34

-------
               TABLE 7.  TRACE METALS IN DUSTFALL BY AREA*
                    No. of months
                      observed
Minimum
Maximum
Mean
Cadmium
Riverhead
Queens
Bronx
Chromium
Riverhead
Queens
Bronx
Copper
Riverhead
Queens
Bronx
Lead
Riverhead
Queens
Bronx
Manganese
Riverhead
Queens
Bronx
Nickel
Riverhead
Queens
Bronx

20
22
22

4
5
7

8
8
8

20
22
22

8
8
8

8
8
8

0.0
0.01
0.0

0.02
0.0
0.06

0.32
0.69
1.85

0.30
1.10
2.20

0.05
0.53
0.68

0.0
0.03
0.05

0.12
0.23
0.22

0.11
0.38
0.60

13.41
11.12
40.68

4.38
23.28
30.99

0.70
2.11
4.57

0.96
0.78
1.45

0.047
0.095
0.103

0.048
0.156
0.371

3.338
3.216
19.009

2.005
10.119
16.367

0.346
1.179
2.860

0.212
0.384
0.850
* Levels in mg/m2/month.
                                 35

-------
                  TABLE 8.  TRACE METALS IN HOUSEDUST BY AREA*

Cadmium
Riverhead
Queens
Bronx
Chromium
Riverhead
Queens
Bronx
Copper
Riverhead
Queens
Bronx
Lead
Riverhead
Queens
Bronx
Manganese
Riverhead
Queens
Bronx
Nickel
Riverhead
Queens
Bronx
Barium
Riverhead
Queens
Bronx

No. of Samples

37
29
33

37
29
33

37
29
33

37
29
33

37
29
33

37
29
33

37
29
33

Minimum

0.1
4.0
4.5

4.0
12.0
28.0

23.8
50.4
72.5

42.6
188.0
124.0

27.9
16.3
37.0

1.0
2.0
18.0

10.0
32.0
75.0

Maximum Geo. Mean

32.0
34.0
118.0

400.0
210.0
230.0

2250.0
900.0
1250.0

1630.0
3180.0
2930.0

158.0
455.0
275.0

95.0
200.0
250.0

430.0
4000.0
13000.0


7.6
13.3
14.6

31.7
42.3
45.0

109.3
196.9
233.7

278.9
629.0
766.3

79.9
131.0
153.4

14.6
26.1
42.0

65.2
137.6
312.4
(continued)
* Levels in yg/g
                                  36

-------
TABLE 8.  (CONTINUED)

Boron
Rlverhead
Queens
Bronx
Iron
Riverhead
Queens
Bronx
Lithium
Riverhead
Queens
Bronx
Mercury
Riverhead
Queens
Bronx
Selenium
Riverhead
Queens
Bronx
Silver
Riverhead
Queens
Bronx
Tin
Riverhead
Queens
Bronx
Vanadium
Riverhead
Queens
Bronx
No. of Samples

37
29
33

37
29
33

37
29
33

37
29
34

37
29
33

37
29
33

37
29
33

37
29
33
Minimum

1.0
2.0
4.0

2320.0
6650.0
2500.0

0.9
2.5
1.3

0.4
0.5
0.5

0.005
0.005
0.005

0.1
0.1
0.1

2.0
2.0
2.0

1.0
1.0
3.0
Maximum

110.0
370.0
410.0

16900.0
92400.0
41000.0

5.8
8.7
10.1

9.1
116.0
19.8

0.255
0.234
0.252

7.0
7.0
38.0

270.0
76.0
230.0

170.0
77.0
90.0
Geo. Mean

14.1
28.5
30.8

6091.6
11195.4
11431.2

3.0
4.3
4.9

1.9
5.9
3.6

0.038
0.062
0.049

0.4
0.5
0.8

6.4
6.2
7.8

14.3
18.8
40.2
                                 37

-------
                   TABLE 9.  TRACE METALS IN SOIL BY AREA*

Cadmium
Riverhead
Queens
Bronx
Copper
Riverhead
Queens
Bronx
Lead
Riverhead
Queens
Bronx
Manganese
Riverhead
Queens
Bronx
Nickel
Riverhead
Queens
Bronx
No. of Samples

56
56
60

56
56
60

56
56
60

56
56
60

56
56
60
Minimum

0.04
0.20
0.11

0.8
6.9
3.9

4.8
33.2
37.0

11.8
28.8
91.5

0.4
1.2
2.5
Maximum

1.72
8.75
5.82

28.0
1020.0
405.0

407.0
1010.0
1660.0

132.0
269.0
283.0

2.7
37.4
25.2
Geo. Mean

0.27
0.89
1.28

5.8
29.3
37.4

32.6
150.5
298.2

34.6
113.1
161.0

1.2
4.3
8.4
*Levels in pg/g,
                                 38

-------
    TABLE 10.  CORRELATION COEFFICIENTS OF LOGS OF MEDIA TRACE METAL LEVELS*

Cadmi urn
Chromium
Copper
Lead
Manganese
Nickel
Zinc
Dustfall-Housedust
(N=99)
0.37
0.23
0.30
0.56
0.53
0.40
0.51
Dustf all -Soil
(N=172)
0.64
-
0.43
0.73
0.82
0.80
0.72
Housedust-Soil
(N=36)
0.25
-
0.52
0.77
0.54
0.41
0.56
* All  of the above correlation coefficients  are  significant at  p  <0.005  except
  for chromium dustfall-housedust,  with  p =  0.02.
                                       39

-------
             TABLE  11.  TESTS  OF  SIGNIFICANCE OF THE EFFECT OF
                        SELECTED  FACTORS ON SOIL TRACE METAL LEVELS*

Cadmium
Copper
Lead
Manganese
Nickel
Zinc
Dustfall
Levels
<0.0001
<0.0001
<0.0001
<0.0001
<0.0001
<0.0001
Homes Within
Areas
<0.0001
0.28
<0.0001
<0.0001
0.0006
<0.0001
Front Yard-Back Yard
Differences
.0004
.99
.0059
.0162
.0029
.0705
* Values given are probabilities of no relationship of factors to soil
  trace metal levels given the observed differences in soil trace metal
  levels between factor categories.  Logs of soil and dustfall levels
  were used in the analysis.
                                    40

-------
          TABLE 12.  TESTS OF SIGNIFICANCE OF THE EFFECT OF SELECTED  FACTORS ON  SCALP HAIR TRACE
                     METAL LEVELS, USING (1) DUSTFALL AND (2)  HOUSEDUST AS A MEASURE OF ENVIRONMENTAL EXPOSURE
(1) Dustfall


Metal
Cd
Cu
Cr
Pb
Mn
Ni

Dust-
fall
_
_
.006
.005
_
.02


Age
—
_
_
.005
.02

Children

Sex Educ.
.0003
.0001
_
.003
.0001
.0001

Hair Haircut
Color Location
— —
,
_
_
_ _
.04

Dust-
fall
—
-
_
.01
_
~


Age
—
.005
_
_
_
m


Sex
.0001
.0001
-
.02
.0001
.0001
Adults
Hair Haircut
Educ. Color Location
.004
.
.
.02
-
- •*


Smoking
_
-
_
.06
.
•
(2) Housedust


Metal
Cd
Cu
Cr
Pb
Mn
Mi
Ba
B
Fe
Li
Hg
Se
Ag
Sn
V

House-
dust
—
_
_
.005
-
.06
.0001
—
.
-
.0001
Mo Test
_
.0001
.0001


Age
—
—
—
.005
.03;
—
.ooor

—
•
—
.
_
—
-
Children

Sex Educ.
.006
.0001

.003
.0001
.0001
.0001

.004
.01
.01

.0002
.0003
.0003

Hair Haircut
Color Location
— —
_
_ _
_
_ _
.04
.03
• ••
_ _
-
.06
.02
_
_
.03

House-
dust
_
_
_
.01
-
.
.0006
_
_
-
.04
No Test
-
-
.0001


Age
_
.004
_
.
-
-
.01
_
_
-
-
-
-
.0001
-


Sex
.0002
.0001
_
.02
.0001
.0001
.0001
.
.
.003
.001
-
-
-
—
Adults
Hair Haircut
Educ. Color Location
.005
_
.
.02
.
_
.05
.04
.04
...
_
_
-
-
.03


Smoking
-
-
-
.06
-
-
-
-
.04
-
-
.02
-
-
-
Values given are the probability of the observed difference in sample mean levels between factor categories assuming
no difference in the original population.  Only values of .06 or less are listed.

-------
TABLE 13.  DUSTFALL, HOUSEDUST, AND SCALP HAIR MEAN TRACE METAL
           LEVELS BY COMMUNITY FOR SCALP HAIR METALS WITH A
           SIGNIFICANT ENVIRONMENTAL EXPOSURE EFFECT

Chromium
Riverhead
Queens
Bronx
Lead
Riverhead
Queens
Bronx
Nickel
Riverhead
Queens
Bronx
Barium
Riverhead
Queens
Bronx
Mercury
Riverhead
Queens
Bronx
Tin
Riverhead
Queens
Bronx
Vanadium
Riverhead
Queens
Bronx
Mean
Dustfall
(mg/m2/mo)
0.05
0.16
0.37

2.01
10.12
16.37

0.21
0.38
0.85

_

—

-

-
Geometric
Mean
Housedust
(ug/g)
31.7
42.3
45.0

278.9
629.0
766.3

14.6
26.1
42.0

65.2
137.6
312.4

1.9
5.9
3.6

6.4
6.2
7.8

14.3
18.8
40.2
Childrens1
Geometric Mean
Scalp Hair
(pg/g)
0.52
0.45
0.80

11.74
14.07
17.13

0.50
0.40
0.70

0.64
0.63
1.26

0.48
1.00
0.73

0.54
0.44
0.83

0.20
0.24
0.40
Adults'
Geometric Mean
Scalp Hair
(ug/g)
-

10.39
14.50
12.88

-

1.11
1.36
2.34

0.66
1.01
0.67

-

0.12
0.20
0.35
                            42

-------
TABLE 14.  GEOMETRIC MEAN SCALP HAIR TRACE METAL CONCENTRATIONS
           BY SEX FOR CHILDREN AND ADULTS*
Children
Metal
Cd
Cu
Cr*
Pb
Mn
N1
Ba
B*
Fe*
Li
Hg
Se*
Ag
Sn
V*
Males
0.77
9.92
0.52
11.86
0.43
0.39
0.54
0.81
18.9
0.04
0.59
0.31
0.18
0.47
0.20
Females
1.14
17.91
0.62
17.52
0.88
0.79
1.30
1.00
23.4
0.05
0.84
0.34
0.28
0.77
0.37
Adults
Males
0.96
13.87
0.57
13.95
0.64
0.47
0.82
0.90
19.8
0.04
0.58
0.35
0.18
0.54
0.18
Females
0.62
25.06
0.63
10.97
1.34
1.14
2.41
1.04
24.0
0.05
0.99
0.27
0.15
1.17
0.19
* No differences by sex for these metals at the 0.05 level  of
  significance.

* Levels given in
                               43

-------
                                REFERENCES


 1.  Hammer, D. I., Finklea, J. F., Hendricks,  R.  H.,  and  Shy, C. M., Hair
     trace metal  levels and environmental  exposure.  Am. 0.  Epidemiol.
     93, 84, 1971.

 2.  Hammer, D. I., Finklea, J. F., Hendricks,  R.  H.,  Hinners, T. A.,
     Riggan, W. B., and Shy, C. M., Trace metals in  human  hair as a
     simple epidemiologic monitor of environmental exposure.  In Trace
     Substances in Environmental Health - V.   D. D.  Hemphill, Ed. Uni-
     versity of Missouri Press, Columbia, Mo.,  1972, pp  25-38.

 3.  Klevay, L. M., Hair as a biopsy material:   III.   Assessment of
     environmental lead exposure.  Arch. Environ.  Health 26, 169, 1973.

 4.  Kopito, L.,  Byers, R. K., and Shwachman, H.,  Lead in  hair of children
     with chronic lead poisoning.  New Eng. J.  Med.  276, 949, 1967.

 5.  Yamaguchi, S., Matsumoto, H., Matsuo, S.,  Kaku, S., and Hoshide, M.,
     Relationship between mercury content of hair and  amount of fish
     consumed.  HSMHA Health Reports 86_, 904, 1971.

 6.  Nelson, N. et al., Hazards of mercury.  Environ.  Res.  4_, 20, 1971.

 7.  Riggan, W. B., Hammer, D. I., Finklea, J.  F., Hasselblad, V., Sharp,
     C. R., Burton, R. M., and Shy, C. M., CHESS,  A  Community Health  and
     Environmental Surveillance System.  In Proceedings  of the Sixth
     Berkeley Symposium on Mathematical Statistics ana Probability
     (Vol. 6), University of California Press,  Berkeley, Calif., 1972.

 8.  Lewis, G. P., Jusko, W. J., Coughlin, L. L.,  and  Hartz, S., Contribution
     of cigarette smoking to cadmium accumulation in man.   The Lancet
     1, 291, 1972.

 9.  National Research Council, Airborne Lead in Perspective.  National
     Academy of Sciences, Washington, D. C., 1972, p.  62,  64, 68.

10.  Pinkerton, C., Creason, J. P., Hammer, D.  I., and Colucci, A. V.,
     Multi-media indices of environmental trace metal  exposure in
     humans.  In Trace Element Metabolism in Animals - 2,  W. Hoekstra,
     J. Suttie, H. Ganther, and W. Mertz, Eds.  University  Park Press,
     Baltimore, Md., 1974, pp. 465-469.

11.  Underwood, E. J., Trace Elements in Human and Animal  Nutrition,  3rd
     ed., Academic Press, New York, N. Y., 1971.

12.  Hinners, T.  A., Kent, J., Terrill, W., Heiderscheit,  T., Burton, R.,
     and Colucci, A. V., Metals in atmospheric particulates collected by
     Hi-Vol sampler and by dustfall bucket.  EPA In-House  Technical
     Report, June 1972.


                                     44

-------
 13.  U. S. Environmental Protection Agency, CHESS measurement methods.
     In Health Consequences of Sulfur Oxides:  A report from CHESS,
     1970-1971.U. S. EPA, Research Triangle Park, N. C., 1974.
 14.  Harrison, W. W., Yurachek, J. P., and Benson, C. A., The determination
     of trace elements in human hair by atomic absorption spectroscopy.
     Clin. Chem. Acta 23, 83, 1969.

 15.  Schroeder, H. A. and Nason, A. P., Trace metals in human hair.  J.
     Invest. Derm. 53_, 71, 1969.

 16.  Gordus, A. A., Maher, C. C., and Bird, 6. C., Human hair as an
     indicator of trace metal environmental exposure.  In Proceedings of
     the First Annual NSF Trace Contaminants Conference. Oak Ridge
     National Laboratory, 1973.

 17.  Gordus, A. A., Environmental aspects of trace metals on human hair,
     report no. 2.  NSF Grant No. GI-35116, 1973.

 18.  Petering, H. G., Yeager, D. W., and Witherup, S. 0., Trace metal
     content of hair:  II. Cadmium and lead of human hair in relation to
     age and sex.  Arch.  Environ. Health 27., 327, 1973.

 19.  Creason, J. P., McNulty, 0., Heiderscheit, L. T., Swanson, D. H.,  and
     Buechley, R. W., Roadside gradients in atmospheric concentrations  of
     cadmium, lead and zinc.   In Trace Substances in Environmental Health -
     V.  D. D. Hemphill,  Ed.  University of Missouri Press, Columbia, Mo.,
     T972, p. 129.

20.  Lagerwerff, J. V., and Sprecht, A. W., Contamination of roadside soil
     and vegetation with  cadmium, lead, and zinc.  Environ. Sci. and Tech.
     4, 583, 1971.

21.  Hinners, T. A., Kent, J. L., and Terrill, W. J., Soil metal variations.
     Materials Research and Standards (ASTHM), June 1972, p. 46.

22.  Hambidge, K. M., Hambidge, C., Jacobs, M., and Baum, J. D., Low
     levels of zinc in hair,  anorexia, poor growth, and hypogeusia in
     children.  Pediat. Res.  6, 868, 1972.

23.  Bate, L. C., Absorption  and elution of trace elements on human  hair
     Int.  J. Appl. Radio. Isotop. V7_, 417, 1966.

24.  Bate, L. C., The use of  activation analysis 1n procedures for the
     removal and characterization of the surface contaminants of hair.
     J. Foren. Sci. 1J), 60, 1965.

25.  Petering, H. G., Yeager, D. W., and Witherup, S. 0., Trace metal
     content of hair:  I. Zinc and copper content of human hair in
     relation to age and  sex.  Arch. Environ.  Health 23_, 202, 1971.
                                    45

-------
 26.   Klevay,  L.  M.,  Hair  as  a  biopsy material.  Am. J. Clin. Nutr.  23,
       •3/7,  1970.                                                    ~~~
 27 '  wlyJns 2'if  Fratjamie!  R:  ?•• ""I. 0. A., Pories, W. J., and Strain,
      W. H., Malversation  in hair analyses.  Am. J. Clin. Nutr. 24, 893,
      I •/ / I •
 28'  rnnJ?ldP'+*- M> 'r^*"*1 *"l M' "" ' and Jacobs • M" A« •  Hair chromium
      concentration:  Effects of sample washing and external  environments.
      Am. J. Clin. Nutr. 25_,  384, 1972.

 29.  Eads, E. A., and Lambdin, C. E. , A survey of trace metals in  human
      hair.  Environ. Res. 6_, 247, 1973.
 3°"  anHU|nlk> J"nGiS]aSOn' *?•' Maes' P" ^Millan,  D.  K. ,  D'Auria, J.,
      and Pate, B. D., The variation of trace element  concentrations in
      single human lead hairs.  J. Radioanal. Chem.  ]_5,  115,  1973.

 31.  Gionanoli-Jakubczak, T., Greenwood, M.  R. ,  Smith,  J. C.,  and Clarkson,
      I. w., Determination of total  and inorganic mercury in  hair by
      fameless atomic absoprtion and of methylmercury by gas chromatography.
      1*1 in.  inem.  
-------
40.  Hambidge, K. M., and Baum, J. D.t Hair chromium concentrations of
     human newborn and changes during infancy.  Am. J. Clin.  Nutr.  25,
     376, 1972.                                                    ~~

41.  Hambidge, K. M., Chromium nutrition 1n the mother and the growing
     child.  In Newer Trace Elements in Nutrition. W. Mertz and W.
     Cornatzer, Eds., Marcel Dekker, New York, N. Y., 1971, pp. 171-194.

42.  Reinhold, J. G., Kfoury, G. A., Ghalambor, M. A., and Bennett, J. C.,
     Zinc and copper concentrations in hair of Iranian villagers.   Am.
     J. Clin. Nutr. 18., 294, 1966.

43.  Pories, W. J., Strain, W. H., Rob, C.  G., Henzel, J.  H., Hennessen,
     J. A., and Plecha, F. R., Trace elements and wound healing.   In
     First Annual Conference on Trace Substances in Environmental  Health.
     D. D. Hemphill, Ed. University of Missouri Press, Columbia,  Mo.,  1967,
     pp. 114-133.

44.  Strain, W. H., Pories, W. J., Flynn, A., and Hill, 0. A., Trace
     element nutriture and metabolism through head hair analysis.   In
     Trace Substances in Environmental Health - V.  D. D.  Hemphill, Ed.
     University of Missouri Press, Columbia, Mo., 1972, pp. 383-397.

45.  Klevay, L. M., Hair as a biopsy material:  I. Assessment of zinc
     nutriture.  Am. J. Clin. Nutr. 23_, 284, 1970.

46.  Prasad, A.,  Metabolism of zinc and its deficiency in  human subjects.
     In Zinc Metabolism. A. Prasad, Ed. Thomas, Springfield,  111.,  1966,
     p. Z57J!

47.  Reinhold,  J. G., Kfoury, G. A., and Thomas, T. A., Zinc, copper and
     iron concentrations in hair and other tissues:  Effects  of low
     zinc and low protein intakes in rats.   J. Nutr. 92_, 173, 1967.

48.  Hunt, W. F., Pinkerton, C., McNulty, 0., and Creason, J. P., A
     study in trace element pollution of air in 77 midwestern cities.   In
     Trace Substances in Environmental Health - IV.  D. D. Hemphill, Ed.
     University of Missouri Press, Columbia, Mo., 1970.

49.  Anke, M.,  and Schneider, H. J., untersuchungen "uber  den Mineral -
     stoffgehalt  der Frauen - und Mannerhaare.  Deut. Z. Verdau.
     Stoff.  22_, 31, 1962.

50.  Klevay, L. M., Hair as a biopsy material:  II. Assessment of copper
     nutriture.  Am J.  Clin. Nutr. 23_, 1194, 1970.

51.  Flesch, P.,  Hair growth.   In Physiology and Biochemistry of the Skin.
     S. Rothman,  Ed.  University of Chicago  Press, Chicago, 111.,  1954,
     pp. 601-661.
                                     47

-------
 52.  Schroeder, H. A., Nason, A. P., and Tipton, I. H., Chromium deficiency
      as a factor  in atherosclerosis.  J. Chron. Dis. 23, 123, 1970.

 53.  Schroeder, H. A., Balassa, J. J., and Tipton, I. H., Essential  trace
      ?n  lLin-,oan:  Man9anese-  A study in homeostasis.  J.  Chron.  Dis.
      19, 545, 1966.

 54.  Schroeder, H. A., Nason, A. P., Tipton, I. H., Balassa,  J.  J.,
      Essential trace metals in man:  Copper.  J. Chron. Dis.  J9., 1007,  1966.

 55.  Nordberg, G. F., and Nishiyama, K., Whole-body and hair  retention  of
      cadmium in mice.  Arch. Environ. Health 24_, 209, 1972.

 56.  Becker, W.  M. , and Hoekstra, W. 6., Dialysis studies of  liver zinc
      in zinc deficient and control  rats.  J. Nutr.  94_,  455, 1968.

 57.  Strain, W  H., Berliner, W. P., Lankau, C. A., McEvoy, R. K. , Pories,
      W. J.,  and  Greerilaw, R. H. , Retention of radioisotopes by hair,  bone
      and vascular tissue.  J.  Nucl. Med. 5_, 664, 1964.
 59.   Beritie,  T.,  Lead concentration found in human  blood  in association
      with lead colic.   Arch.  Environ.  Health 23,  289,  1971.

 60.   Piscator, M., and Axelsson,  B. , Serum proteins  and  kidney function
      after exposure to cadmium.   Arch.  Environ. Health 21_, 604, 1970.

 61.   Winge,  D.  R., and Rojagopalan,  K.  V.,  Purification  and some
      properties of Cd-binding protein  from rat liver.  Arch. Biochem.
      Biophy. 153.  755, 1972.

 62.   Parizek,  J.,  Ostadalova, I.,  Kalouskova,  J., Babicky, A., and Benes,
      J.,  The detoxifying  effects  of  selenium interreations between
      compounds  of  selenium and certain  metals.  In Newer Trace Elements in
      Nutrition.  W.  Mertz  and  W. Cornatzer,  Eds., Marcel Dekker, New Vork,
      N. Y. , 1 971 ,  p. 96.

 63.   Mertz, W.,  and Roginski, E.  E., Chromium  Metabolism:  The glucose
      tolerance  factor.  In Newer Trace  Elements in Nutrition. W.  Mertz
      and  W. Cornatzer,  Eds. Marcel Dekker,  New York, N. Y., 1971, pp. 134-
      1 35.

64.   Hambridge,  K.  M.,  Chromium nutrition in the mother and the growing
      child.  In  Newer  Trace Elements in Nutrition. W. Mertz and W.
      Cornatzer,  Eds. Marcel Dekker,  New Vork,  N. V., 1971, pp.  169-194.

65.   Hopkins, L. L., and Mohr, H. E., The biological essentiality of
      vanadium.    In  Newer Trace Elements in  Nutrition. W.  Mertz  and
     W. Cornatzer,  Eds. Marcel Dekker, New  York, N.  Y., 1971, pp.  203-204.


                                    48

-------
66.  McBean, L.  D.,  Mahloudji, M.,  Reinhold, J. G., and Halsted, J. A.,
     Correlation of  zinc  concentrations  in  human  plasma and hair.
     Am. J.  Clin. Nutr. 24,  506,  1971.

67.  Hambidge, K. M.,  and Rogerson, D. 0.,  Comparison of hair chromium
     levels of nulUparous and parous women.   Am.  J. Obstet. Gynec.
     103, 320, 1969.

68.  Strain, W.  J.,  Steadman, L.  T., Lankau, C. A., Berliner, W. P. and
     Pories, W.  J.,  Analysis of  zinc levels in hair for the diagnosis
     of zinc deficiency in man.   J. Lab. Clin. Med. 68., 244, 1966.

69.  Piscator, M., and Lind, B.,  Cadmium, zinc, copper, and lead in
     human renal cortex.   Arch.  Environ. Health 24_, 426,  1972.

70.  Hammer, D.  I.,  Colucci, A.  V., Hasselblad, V., Williams, M. E.,
     and Pinkerton,  C., Cadmium  and lead in autopsy tissues.  J. Occup.
     Med. 15., 956, 1973.

71.  Schroeder, H. A., and Tipton, I.  H., The  human body  burden of lead.
     Arch. Environ.  Health 17.,  965, 1968.

72.  National Research Council,  Vanadium. National Academy of Sciences,
     Washington, D.  C., 1974, p.TT
                                    49

-------
                                   TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing)
 1  REPORT NO
   EPA-600/1-78-037a	
 4 TITLE AND SUBTITLE
   HUMAN SCALP HAIR: AN  ENVIRONMENTAL EXPOSURE INDEX FOR
   TRACE ELEMENTS.  I.   Fifteen Trace Elements in
   New York, N.Y.  (1971-72)
                                                           3 RECIPIENT'S ACCESSION-NO.
                                                           5 REPORT DATE
               May iQ-
              . PERTORMI
MING ORGANIZATION CODE
 7 AUTHOR(S)
                                                           8. PERFORMING ORGANIZATION REPORT NO
   John P. Creason, Thomas  A.  Hinners, Joseph E. Bumgarne
   and Cecil Pinkerton
 9 PERFORMING ORGANIZATION NAME AND ADDRESS
   Health Effects Research  Laboratory and Environmental
   Monitoring and Support  Laboratory
   Office of Research  and  Development
   Research Triangle Park,  N.C.  27711	
             10. PROGRAM ELEMENT NO.

                1AA601
             11. CONTRACT/GRANT NO.
 12 SPONSORING AGENCY NAME AND ADDRESS
   Health Effects Research  Laboratory
   Office of Research  and  Development
   U.S.  Environmental  Protection Agency
   Research Triangle Park.  N.C.  27711
             13 TYPE OF REPORT AND PERIOD COVERED
 RTP.NC
             14 SPONSORING AGENCY CODE
               EPA 600/11
 15 SUPPLEMF.NTARY NOTES
 16 ABSTHACT
        Previous studies  have revealed that hair trace element concentrations can
   reflect exposure  in  cases  of frank poisoning and deficiency.  Correlations have
   been found also in some  populations living in regions where metallurgic processes
   are conducted.

        This study reports  significant correlations between hair barium, chromium,  lead,
   mercury, nickel,  tin,  and  vanadium content and exposures (as measured by analyses
   for the corresponding  elements in dustfall or housedust) within a single
   metropolitan area.   Age,  sex, hair color, and smoking habits were included in the
   statistical evaluation.   Several metals showed a tendency to increase and decrease
   together in the hair specimens in agreement with trends reported for other human
   tissues.
        It is acknowledged  that hair has the capacity to adsorb and to release trace
   elements in certain  situations.  However, population studies can compensate for
   confounding influences by  (1) a randomizing effect, by (2) an averaging effect,  and
   (3) by statistical rejection of unrepresentative data values.  The relationship  of
   hair content to (a)  content in other tissues and to (b) metabolic status are
   separate and complex issues that 'should not be confused with (c) exposure relation-
   c h -i r\c	
                                KF Y WORDS AND DOCUMENT ANALYSIS
                  DESCHIHTORF,
   trace elements
   hair
   indexes (ratios)
   environmental surveys
                                              l> IDENTIFIERS/OPEN ENDED TERMS
                          C COSATI I'lcld/CfOUp
                            06,  T,  F
   UIS TRIBUTION STATEMENT

   RELEASE TO PUBLIC
19 SECURITY CLASS (This Report)
  UNCLASSIFIED
                          21 NO. OF PAGES
        59
                                              20
                                                            Tins page)
                                                                        22 PRICE
EPA Form 2220-1 (9-73)
                                            50

-------