Monitoring
of Persistent Organic Pollutants
in human milk and blood
Anton Kočan, Jana Klánová
Matrices selected for the GMP
 The COP has decided that the air monitoring and
human exposure through breast milk or maternal
blood will be used as core media for the first
evaluation planned in 2009.
 For future evaluations, the COP has also decided
to endeavour to supplement the core data with
data from other media such as biota, water, soil,
and sediments.
A. Kočan, Slovak Medical University
What is Human Biomonitoring (HBM) ? (1)
 HBM is the assessment of human exposure to
chemicals present in air, water, soil, dust, food or other
environmental media via measurement of the chemicals
or chemical metabolites present in human specimens,
such as blood, urine, hair, saliva, feces, tissues. The
result of such measurement is usually called „body
burden“.
 HBM is a tool to help better understand human
exposure to environmental chemicals – both natural
and man-made.
 If gathered from a representative sample of a
population – for instance, children or adults in a
particular area – HBM can be used to document
whether that subgroup as a whole has been exposed to
some chemicals and to what extent. A. Kočan, Slovak Medical University
What is Human Biomonitoring (HBM) ? (2)
 HBM is a scientific technique for assessing human
exposures to environmental agents and their effects,
based on sampling and analysis of an individual´s tissues
and fluids. While blood, urine, breast milk and expelled air
are most commonly measured, hair, nails, fat, bone and
other tissues may also be sampled.
 This technique takes advantage of the knowledge that
environmental agents that have entered the human body
leave markers reflecting this exposure. The marker may
be the agent itself or a breakdown product, but it may also
be some change in the body resulting from the interaction
of the agent or its breakdown product(s) with the
individual, such as alterations in the levels of certain
enzymes or other proteins which may lead to
modifications of normal body processes.
A. Kočan, Slovak Medical University
HBM can be used to :
 Reinforce regulatory actions by providing actual
data about which chemicals get into people and
at what levels.
 Improve exposure assessment.
 Establish baselines or reference ranges.
A. Kočan, Slovak Medical University
 Facilitate people's right to know what chemicals
are in their bodies.
 Establish priorities for tackling environmental
health-related problems.
Some variables affecting HBM (2)
 The precision of chemical analysis is generally believed to
constitute only a minor part of the total variance in monitoring
time-series of environmental data where sample variation is
expected to be large, much larger compared to laboratory
precision.
 That is true if the same accredited laboratory is used through the
whole series. However, if, from year to year, different laboratories
carry out the analysis, it could seriously decrease or disable the
possibility to evaluate time-series of, for example, POPs.
A. Kočan, Slovak Medical University
 The same is true if the same laboratory changes its methodology
and, for example, co-elutions are resolved leading to a decrease in
estimated concentrations unless measures are taken to
compensate for them.
 If detection limits are improved, i.e. analytes are now found where
they were not detected before, that may lead to similar problems
depending on how results below the limit of quantification (LOQ)
are treated.
Some variables affecting HBM (1)
 Fat content and composition in human milk changes
dramatically during the first weeks after birth, which leads
to variation also in analysed POPs. In order to reduce
random variation, sampling should preferably be carried
out during a well defined period, e.g. 3 weeks after birth.
Also the fat content varies considerably depending on
whether sampling is carried out in the beginning or at the
end of the feeding session.
 The use of narrow sampling unit definition implies that a
smaller part of the studied population is represented.
Often, this leads to unfounded assumptions of similar
trends, e.g., for both sexes or for various age classes. To
improve representativity, if economy permits, stratified
sampling should be applied rather than allowing for a
wider definition of the sampling unit.
A. Kočan, Slovak Medical University
Ethical issues relevant to HBM
 Ethical issues are relevant in almost every facet of human
biomarker research from the design of the studies, the
identification and recruitment of subjects, to the handling
and use of the data, and interpretation and
communication of the results.
 The right to monitor chemical pollutants in blood and
breast milk is a crucial aspect of community right to know
but also brings with it responsibilities to care and support
those who are tested. .
 Researchers and regulators have to be aware of the
potential for biomarker information to affect the lives of
subjects and their families. There must be sufficient
protection of personally identifiable data and regulation of
its use, while ensuring individual subjects right to know
their results.
A. Kočan, Slovak Medical University
Number of samples needed
 More samples provide more precise and reliable
estimates of mean concentrations and variance.
However, the contributions from additional samples
depend to a very high degree on the sampling
strategy.
A. Kočan, Slovak Medical University
 Using pooled samples of several specimens will
decrease the number of chemical analyses required
to estimate a reliable mean concentrations compared
to one or a few individual samples, since a larger
proportion of the total population is represented.
Disadvantages with pooled samples are that extreme
values from single specimens may influence the
concentration of the pool without being revealed.
 All sampling should follow established methodological
guidelines, which should be agreed upon before the start of
any programme activity in a region.
 If possible, samples in all programmes should be numbered in
the same way.
A. Kočan, Slovak Medical University
General considerations pertaining to GMP sampling (1)
 The sampling window for the initial baseline will be 2003 ± 5
years.
 Sample frequency and timing should, as much as possible, be
harmonized between matrices.
As a rule, samples should be taken at least annually and during
the same period every year.
For some matrices where seasonal influences would be less
important (e.g. human breast milk), the sampling frequency and
duration might be different.
For the statistical analysis of the levels it would be preferable
to take many samples frequently from one location rather than
to take a few samples from many different locations.
A. Kočan, Slovak Medical University
General considerations pertaining to GMP sampling (2)
 Sampling should always include field or trip blanks and, to the
extent possible, duplicate samples for the purpose of sample
sharing and the analysis of variance.
 Sample banking should be considered for all samples. Sample
banking is an expensive and resource intensive activity that
needs to be sustainable in a long time perspective. However, if
properly managed it may yield important insights into
exposures over time (e.g for new POPs) and may also be used
for retrospective studies.
A sample bank at the Slovak Medical University in Bratislava
Human milk and maternal blood as matrices for the GMP (1)
 Human milk and maternal blood are both good sample media for
assessing POPs exposure in humans.
Furthermore, both these media can be used to demonstrate
possible temporal trends and regional variations in levels.
 Human milk sampling is non-invasive and milk can generally be
obtained from lactating mothers in reasonable quantities. In
certain populations it may, however, be difficult to obtain human
milk samples in the required time period, i.e., 2 to 4 weeks after
delivery.
 Blood sampling is invasive, but sampling of mothers prior to
giving birth may readily be achieved. However, blood sampling
may not be acceptable in certain cultures.
A. Kočan, Slovak Medical University
 Depending on local considerations, biological samples of human
origin, including blood and milk, should be considered a potential
biohazard. Necessary precaution procedures should be applied to
both sampling and handling of all samples, not only in situations
where one may expect a problem, e.g. HIV-positive serology and
hepatitis.
Human milk and maternal blood as matrices for the GMP (2)
 The limit of detection for POPs will in general be lower in
milk than in blood.
The reason for this is partly the difference in lipids
between the media and the fact that larger volumes of milk
as compared to blood can normally be obtained.
When the limit of detection is approached the analytical
precision will decrease.
 An important consideration in the choice of human milk
and maternal blood as biological indicators is that we will
only obtain information from a specific part of the
population both with regard to gender and age.
As for future trend studies, a careful evaluation should be
done to explore alternative representative groups in a
population, e.g. men (specified age groups), youth groups
of both gender, school children or infants.
A. Kočan, Slovak Medical University
Human milk and maternal blood as matrices for the GMP (3)
 A population study must be based on sampling and analyses of
individual samples; human milk or maternal blood.
Pooled samples might be considered for certain contaminants, such
as the dioxins which are expensive to analyze and need larger sample
volumes.
 In order to reduce sample variance and facilitate comparability a
stratified sample design should be adopted.
This should be based on demographic information collected in
specific questionnaires, i.e. age, residence, occupational history,
smoking habits, current and previous diet, etc.
A. Kočan, Slovak Medical University
 Selection of study groups should be based on known exposure
patterns, global or local.
The groups with known high exposure levels are more sensitive to
changes in the environment and will provide better indications in
trend analyses. Even in countries with very limited background
information one might be able to select population groups of interest,
such as rural versus urban; fish eating populations versus rural
agricultural populations with high exposures to pesticides;
populations living in areas with re-introduction of DDT for malaria
prophylaxis etc.
Human milk and maternal blood as matrices for the GMP (4)
 Sample size will depend on the circumstances, and to estimate the
number of samples needed a number of factors have to be considered
to achieve representative samples. For either human milk or blood, 50
individual samples are to be collected. However, new technologies
and new, certified laboratories will provide the opportunity to begin
epidemiological studies with individual results on a larger scale.
 The choice of milk or blood depends very much on the practical
implementation regionally or locally. Two examples:
A. Kočan, Slovak Medical University
 In the Arctic many indigenous women deliver their babies and go
home to the tundra before they have started their milk production.
To collect colostrum provides a very different medium than the
fully developed breast milk 2-3 weeks after delivery. It is not
possible to trace the women at the right time for breast milk
collection. A blood sample will solve that problem.
 In certain areas of Africa sampling of maternal blood might be
problematic. In those cases breast milk is the best matrix.
 Trained personnel is crucial at the sampling and analytical stages.
Standardized protocols, equipment and education of field personnel
as well as laboratory personnel must be implemented.
Human milk and maternal blood sampling
 Human milk should be sampled according to a WHO protocol.
 At least 50 individual milk (blood) samples should be collected. But
countries with populations over 50 million should include at least one
additional participant per one million population over 50 million.
Countries with populations well over 50 million (or with sufficient
resources) are encouraged to prepare a second pooled sample (or
more) if feasible.
A. Kočan, Slovak Medical University
 Mother should be primiparae
 Mothers who may have unusually high exposure to POPs (living in the
vicinity of incinerators, pulp or metal industiries or where organochlorine
chemicals have been produced) should not be included in the survey to
avoid skewing the results.
 Maternal blood should be sampled according to an AMAP protocol.
 Selection criteria for mothers:
 Mother should be under 30 years of age
 Both mother and child should be apparently healthy, including
normal pregnancy
 Mother should be breastfeeding one child only (i.e., no twins)
 Mother should have resided in the area for at least the previous 10 years
How to identify possible donors (1)
A. Kočan, Slovak Medical University
 Possible donors can be contacted before giving birth In countries
with adequate pre-natal coverage.
 Selection before giving birth:
 All potential donors should be informed about the benefits of
breastfeeding and be encouraged to breastfeed even if they do not
intend to or are not selected to participate in the survey.
 Once a participant indicates a willingness to take part in the survey,
she should be invited to complete certain section of the
questionnaire.
The questionnaire can be completed through a personal interview
at the pre-natal clinic or completed by the potential donor at home
and returned to the clinic, either in person or by mail.
 Depending on the homogeneity of the population, up to 250
completed questionnaires should be collected. Depending on the
country, more than 50 potential and 10 reserve donors should be
selected to take into account possible withdrawals. Participants
should be notified of their selection and where and when the
sample will be collected.
How to identify possible donors (2)
A. Kočan, Slovak Medical University
 Samples are collected of mothers at postnatal clinics and other
venues, e.g., well-baby clinics without pre-selection.
 Selection after giving birth:
 Mothers can be contacted directly and interviewed to complete all
section of the questionnaire.
 Samples can then be collected immediately or at another
appropriate time.
 While this method can reduce the time of the survey by up to 3
months, it does not allow for further stratification of the cohort to
reduce variability.
 However, after the cohort selection criteria have been established
from the first sample collection, this method offer advantages for
the second and subsequent sample collections.
Collection of samples
A. Kočan, Slovak Medical University
 Sampling can be carried out between 3 to 8 weeks after delivery.
 At least 50 ml of milk in total should be collected by hand expression after a
feeding or while infant is nursing on the other breast, to take advantage of the
let-down reflex of the mother.
A human milk pump to facilitate expression can be applied.
If necessary, the mother may collect the sample at home, in which case
manual expression is preferred. If so, she should be given detailed
instructions for sampling, storing and transporting of milk samples.
Mothers should also be given a clean glass jar with a protected screw cap to
collect and store the milk sample.
 The sample should be collected directly to the collecting jar and, if collected
at home, stored in the collecting jar in the home freezer until it can be
delivered.
Otherwise milk samples may be stored in the refrigerator at about 4 °C for a
maximum of 72 hours.
If refrigeration is not available, a tablet of K2Cr2O7 may be added to chemically
sterilize the milk. If the milk is to be collected at home, the tablet may be
placed in the collection jar before it is given to the donor.
CAUTION: The mother must be told to keep the jar with potassium dichromate away
from other children in the household as this is a toxic chemical.
Questionnaire for mothers donating breast milk (1)
Questionnaire for mothers donating breast milk (2)
• SPE (10 g C-18 column), 10 - 30 mL serum
PCDD/PCDF/PCB analysis (1)
C-18 column
Blood serum
• SPE (10 g C-18 column), 10 - 30 mL serum
• Clean-up using semi-automated equipment
(H2SO4/silica, alumina, carbon chromatography) :
PCDD/PCDF/PCB analysis (1)
Semi-automated sample cleanup system
Hexane eluate
from an SPE column
Silica/H2SO4
layer
Silica/KOH
layer
Alumina column Carbon column
• SPE (10 g C-18 column), 10 - 30 mL serum
• Clean-up using semi-automated equipment
(H2SO4/silica, alumina, carbon chromatography) :
• Fraction 1:
• Fraction 2:
• WHO mono-ortho-PCBs (105, 114, 118, 123, 156, 157,
167, 189)
• other ortho PCBs (18, 28, 44, 49, 52, 66, 74, 87, 99, 101,
110, 128, 138, 146, 149, 151, 153, 170, 172, 177, 178, 180,
183, 187, 194, 195, 196/203, 201, 206, 209)
• Seven 2378-PCDDs
• Ten 2378-PCDFs
• Four coplanar PCBs (77, 81, 126, 169)
PCDD/PCDF/PCB analysis (1)
PCDD/PCDF/PCB analysis (2)
• 1 analysis batch is composed of:
• 14 serum samples
• 1 in-house RM (spiked porcine serum)
• 1 blank (no serum)
• occasionally CRM (serum)
• HRGC (DB5-ms) separation
• HRMS (10000 res) quantification
• Isotope dilution method based on EPA 1613
and 1668
WHO-TEQpg/glipidwt
0
2
4
6
8
10
12
14
16
18
20
PCDDs/PCDFs
PCBs
European comparison of PCDD/F and PCB levels
in Human milk (WHO Exposure Study 2001/2002)
Source: Van Leuwen, R. Malish 2002 "WHO exposure study on the levels of PCBs, PCDDs and PCDFs in Human Milk" 3rd round A. Kočan, Slovak Medical University
0
5
10
15
20
25
30
35
40
ppt,lipidwt
Time trends in PCDD/F levels in human milk
(WHO Exposure Studies)
Source: WHO A. Kočan, Slovak Medical University
1988 1993 2002 2007
Source: Van Leuwen, R. Malish 2002 "WHO exposure study on the levels of PCBs, PCDDs and PCDFs in Human Milk" 3rd round A. Kočan, Slovak Medical University
Slovakia‘s districts chosen for exposure studies
A. Kočan, Slovak Medical University
Time trends in PCB levels in human blood
(2 districts in Slovakia)
A. Kočan, Slovak Medical University
0
500
1000
1500
2000
2500
3000
3500
4000
4500
PCBs HCB pp´-DDE
ng/g(lipidadjusted)
MI 1997 (215 blood samples)
MI 2001 (1011 blood samples)
SP 1997 (205 blood samples)
SP/SV 2001 (1038 blood samples)
“Cons” - Why is Public Health Biomonitoring
Controversial?
 May cause unfounded fear. Because it can be
measured, doesn’t mean it causes harm.
 Hard to tell how much is too much. No
“meaning” without reference ranges.
 Provides no information about sources.
A. Kočan, Slovak Medical University
 May modify behavior in negative ways.
 May cause valuable products to be banned.
“Pros” - Biomonitoring helps answer
Important Public Health Questions
 What are we exposed to and how much?
 Do public health policies and regulatory
programs reduce exposures over time?
 Are some groups more highly exposed than the
general population?
A. Kočan, Slovak Medical University
 What are the relationships between exposure and
disease?
Source: P. Flessel, EHL, California DHS
Why to breastfeed
 Nutrition:
A. Kočan, Slovak Medical University
 Breast milk provides, in an easily digested form, all the nutrients an
infant needs for the first six months of life. Exclusive breastfeeding (i.e.,
no other food or drinks given, not even water) for the first six months
offers maximum protection to infants against pneumonia, diarrhoea and
other common infections of childhood.
 Up to 2 years of age or more, breast milk continues to provide high-quality
nutrients and helps protect against infection.
From 6 to 12 months, breast milk usually provides 60–80% of all energy,
protein and other nutritional requirements – e.g., vitamins and other
micronutrients.
From 12 to 23 months, breastfeeding can provide up to 35–40% of these
requirements.
 Psychosocial development:
 Family planning / child spacing:
 Breastfeeding delays the return of a woman’s fertility. A woman who does
not breastfeed is at increased risk of becoming pregnant again as early as
six weeks after the birth of the child.
 Breastfeeding promotes the emotional relationship, or bonding, between
mother and child.
Slovakia‘s Districts Chosen for POPs Exposure Studies:
Nitra
Bratislava
Myjava
Trebišov
Michalovce
Veľký Krtíš
in 1992
- 94
in 1997
in 2001
&
2003-05
A. Kočan, Slovak Medical University
A. Kočan, Slovak Medical University
Michalovce District
Stropkov District
PCB Concentrations in Ambient Air
Michalovce (polluted district) vs Stropkov (control district)
110
110
100
280
1700
1500
200
180
73
80
85
64
Senne
Hazin
Michalovce I
Michalovce II
Vola
Strazske
Stropkov I
Stropkov II
Potocky
Brusnica
Lomné
Turany n/O.
PCB Concentration [pg/m
3
]
A. Kočan, Slovak Medical University
700
5800
400
170
8,7
6,3
1,6
Near a Chemko waste
storage building
Near a Chemko landfill
site
Fields near a Chemko
landfill site
Rim of the Michalovce
landfill site
Agricultural soil in
Michalovce District (8)
Agricultural soil in
Stropkov District (4)
Soil from the Vihorlat
Hills (2)
ppb (dry soil)
PCB Concentrations in Soil
Vicinity of Chemko disposal sites and agricultural fields
Waste Disposal Sites
Agricultural/Forest Soils
A. Kočan, Slovak Medical University
8,5
33
2400
6000
52
3900
1600
Domasa Lake
Ondava River
Zemplinska Sirava Lake
Filling canal (for ZS)
Laborec / upstream Strazske
Laborec / downstream Strazske
Sewage canal / Strážske town
Chemko effluent canal
ppb (dry sediment)
Michalovce District
Stropkov District
PCB Levels in Sediment Samples
Michalovce (polluted district) vs Stropkov (control district)
3 000 000 = 0.3 %
A. Kočan, Slovak Medical University
81
127
0,6
2,1
Planktonophage+benthic
Predatory
Planktonophage+benthic
Predatory
ppm (sum of 9 congeners, lipid weight)
PCB Levels in Fish
Michalovce (polluted district) vs Stropkov (control district)
Stropkov waters
Michalovce waters
A. Kočan, Slovak Medical University
Average PCB Levels in Eggs and Chicken
Collected in Michalovce and Stropkov Districts
Sum of 28, 52, 101, 118, 138, 153, 156, 170 and 180 congeners
24,5
15,4
18,3
16,2
374,3
128,9
2764,2
Chicken (SP-retail)
Eggs (MI-producer)
Eggs (MI-retail)
Eggs (SP-retail)
Chicken (MI-
retail+home)
Eggs (SP-
homemade)
Eggs (MI-
homemade)
ppb (lipid basis)
A. Kočan, Slovak Medical University
0
20000
40000
60000
80000
100000
15 20 25 30 35 40 45 50 55 60 65 70 75
0,0
20000,0
40000,0
60000,0
80000,0
100000,0
15 20 25 30 35 40 45 50 55 60 65 70 75 Age
Svidnik+Stropkov Michalovce
ng/g, lipid adjusted
PCB Levels - Adults
A. Kočan, Slovak Medical University
0
20000
40000
60000
80000
100000
15 20 25 30 35 40 45 50 55 60 65 70 75
0,0
20000,0
40000,0
60000,0
80000,0
100000,0
15 20 25 30 35 40 45 50 55 60 65 70 75 Age
Svidnik+Stropkov Michalovce
ng/g, lipid adjusted
PCB Levels - Adults
Husband (78)
Wife (72)
both ate local fish
A. Kočan, Slovak Medical University
0
20000
40000
60000
80000
100000
15 20 25 30 35 40 45 50 55 60 65 70 75
0,0
20000,0
40000,0
60000,0
80000,0
100000,0
15 20 25 30 35 40 45 50 55 60 65 70 75 Age
Svidnik+Stropkov Michalovce
ng/g, lipid adjusted
PCB Levels - Adults
Woman (55)
Woman (70)
high portion of food
from home-kept animals
A. Kočan, Slovak Medical University
Median PCB-153, HCB, and p,p’-DDE levels in blood serum vs age
groups (specimens collected within the PCBRISK project in 2001
1665
1056
718
348
151
90
994
367
364
295
212
109
4262
2485
1967
1361
763
473
60 – 78
50 – 59
40 – 49
30 – 39
17 – 29
8 – 9
AgeGroups
PCB-153
HCB
pp'-DDE
ppb, lipid adjusted
N=434
N=369
N=198
N=708
N=683
N=89
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
PCBs PCBs HCB HCB
ppb,bloodlipidadjusted
MI Men
MI Women
SV/SP Men
SV/SP Women
Comparison of POP levels in men and women and time trends
1998
Study
1998
Study
2001
Study
2001
Study
A. Kočan, Slovak Medical University
 32- and 4.3-fold increase of DDT and DDE respectively in a mother before delivery vs the mother 21 months later
 180- and 42-fold increase of DDT and DDE respectively in her 16-m child vs the 6-m child
 35-% decrease of DDT in the child at the age of 21 moths while DDE unchanged
pp'-DDE and pp'-DDT levels in the blood of a mother and her child (#4318)
1020
864
96
4023
4345
4023
67 48 23
4075
2121
2694
0
500
1000
1500
2000
2500
3000
3500
4000
4500
ng/g,lipidadjusted
pp´-DDE
pp´-DDT
Mother‘s and child‘s blood was taken
when the child was 21 months old
A. Kočan, Slovak Medical University
ng/g, lipid adjusted
PCB-153 congener levels in blood serum taken from
mothers and their umbilical cords (141 pairs)
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
1300
0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500
Mothers
Umbilicalcords
R=0.967
A. Kočan, Slovak Medical University
ng/g, lipid adjusted
PCB-153 congener levels in blood serum taken from mothers
and their 6-month-old children (141 pairs)
0
200
400
600
800
1000
1200
1400
1600
1800
2000
2200
2400
0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500
Mothers
6-month-oldchildren
R=0.583
A. Kočan, Slovak Medical University
ng/g, lipid adjusted
PCB-153 congener levels in blood serum taken from children
at the age of 6 and 16 months (141pairs)
0
200
400
600
800
1000
1200
1400
1600
1800
2000
2200
2400
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 2600 2800 3000 3200 3400 3600
16-month-old children
6-moth-oldchildren
R=0.916
A. Kočan, Slovak Medical University
PCB Levels in Human Milk Samples / WHO
Sum of 28, 52, 101, 138, 153, and 180 congeners
0.25
0.49
0.40
0.33
0.49
0.62
1.02
Slovakia 2006
Kosice_MWI 2001
Svidnik 2001
Nitra 2001
Nitra 1993
Michalovce 2001
Michalovce 1993
pg/g, lipid adjusted
A. Kočan, Slovak Medical University
PCDD/F-TEQ Levels in Human Milk Samples / WHO
2.5
3.1
4.2
3.7
4.2
3.5
3.4
4.0
4.8
5.4
4.9
8.5
6.4
11.7
10.1
12.5
12.7
10.7
19.5
Slovakia 2006
Kosice_MWI 2001
Svidnik 2001
Nitra 2001
Nitra 1993
Michalovce 2001
Michalovce 1993
pg/g, lipid adjusted
PCDDs
PCDFs
dl-PCB
A. Kočan, Slovak Medical University
Recommended Literature & Links
Pertaining to Human POPs Monitoring
 Guidance on the Global Monitoring Plan for Persistent Organic Pollutants. UNEP.
February 2007 (preliminary version).
 Guidance for Analysis of Persistent Organic Pollutants (POPs). UNEP. March 2007.
http://www.chem.unep.ch/pops/laboratory/analytical_guidance_en.pdf
 Guidelines for Developing a National Protocol. WHO. March 2007 (revised).
http://www.who.int/foodsafety/chem/POPprotocol.pdf
 http://www.pops.int
 http://www.chem.unep.ch/pops/default.html
 http://www.cdc.gov/biomonitoring
 http://www.aphl.org/programs/environmental_health/biomonitoring/Pages/default.aspx
 http://www.who.int/foodsafety/chem/pops/en/