6/9/2015 Meeting Agenda
Advisory Panel to the
Environmental Health Tracking and Biomonitoring Program
1:00–4:00 pm at The American Lung Association in Minnesota
490 Concordia Avenue, St. Paul, MN
Time
Agenda Items
Presenters
Description/expected outcome
1:00
Welcome & Introductions
Lisa Yost, Chair for
Pat McGovern
Panel members & audience are invited to introduce
themselves.
1:05
2015 Legislative Report
Paul Allwood,
Assistant
Commissioner,
MDH
Information Item
1:15
1:30
East Metro PFC3
Biomonitoring Project
Results Analysis
Christina
Rosebush
Discussion
Paul will update the panel on the 2015 legislative
session and funding for the Environmental Risks
Initiative. Panel members are invited to ask questions
and comment.
Discussion item:
Christina will present the preliminary results from the
PFC3 Project.
Questions for the panel:
Are there panel recommendations for additional
analyses before presenting these results to the public?
What key messages are most important for informing
the community about these findings?
Is presenting the percent change in Geometric Mean
over time or mean change in individual PFC level over
time best for communicating with the public?
Is the inclusion of Washington County residents in the
NHANES subsample concerning? Should MN
Biomonitoring pursue additional information on the
Washington County subsample from NHANES?
1:45
Biomonitoring Updates
Information Item:
Biomonitoring updates are provided in written form.
Panel members are invited to ask questions and
comment on all updates.
•
•
1:50
2:05
MN FEET
East Metro Cancer
Report
State Air and Health
Initiative
Discussion
Jeannette Sample
and
Chuck Stroebel
Discussion item:
Jeannette will present highlights from a new report
that measures the impact of air pollution on health in
the Twin Cities area. Chuck will preview the “Be Air
Aware” website for the public.
Questions for the panel:
•
What key findings from the report are most
important for public communications?
1
Time
Agenda Items
Presenters
Description/expected outcome
•
•
2:25
Refreshments
2:40
Tracking Updates and
Program Evaluation
Matthew
Montesano
Information item:
Jean Johnson,
Blair Sevcik and
Frank Kohlasch,
Section Manager,
Environmental
Analysis &
Outcomes,
MPCA
Discussion item:
•
•
2:55
3:20
Portal Updates
Strategic
Planning
• Upcoming Grant
Year Workplan
Mercury Impact
Analysis for Informing
Reduction Initiatives
Discussion
Matthew will demonstrate the new data visualizer
tool and describe the MN Tracking program
evaluation. Other updates are provided in written
form. Panel members are invited to ask questions and
comment on all updates.
Jean will introduce, and Blair will describe the
methods used in a recent MN Tracking analysis of the
health and economic burden of mercury in newborns.
Frank will describe the Statewide Mercury Reduction
Initiative.
Questions for the panel:
•
•
•
3:50
Public Comments and
Audience Questions
3:55
New Business
4:00
Motion to adjourn
How might the results be used?
What next steps would you recommend?
How can MDH-EHTB continue to support and
inform the MPCA’s mercury reduction initiatives?
Given the limitations, how well does the economic
burden analysis serve its intended purpose?
What additional information would be most
helpful going forward?
Note to audience: The panel asks that audience members hold comments and questions on discussion items
until the end of the meeting, when the chair will invite questions from the audience. Audience members are
asked to identify themselves when they speak, and to please record their names and affiliations on the list at
the sign-in table. Meetings are recorded on audiotape.
2
Table of Contents
2015 Legislative Report…………………………………………………………………………………………………4
East Metro PFC3 Biomonitoring Project Results Analysis ………………………………..…………..5
Biomonitoring Updates …………………………………………………………………………………..…………15
State Air and Health Initiative…………………………………………………………………………..………..21
Tracking Updates and Program Evaluation………………………………………………………..……….24
Mercury Impact Analysis for Informing Reduction Initiatives……………..……………….…..27
Other Information …………………………………………………………………………………………..…………38
3
Section Overview: 2015 Legislative Report
Paul Allwood, Assistant Commissioner for the Minnesota Department of Health, will update the
panel on the 2015 legislative session and the status of the Environmental Risks Initiative
funding (part of the Environment and Natural Resources funding).
This initiative (joint with MPCA) currently funds the MDH state biomonitoring program work
including the PFC3 project in the East Metro community and the MN FEET project (measuring
mercury, lead and cadmium in newborns and pregnant women.) The initiative also funds work
on the joint MPCA/MDH project measuring the health impacts of air pollution in the Twin Cities
metro area, community engagement on air and health risks, and a Health Impact Assessment
(HIA) in a Twin Cities community.
Information Item:
After this presentation, panel members are invited to ask questions and comment.
4
Section Overview: East Metro PFC3 Biomonitoring Project Results
Analysis
Christina Rosebush will present the preliminary results from the PFC3 Project.
Questions for the panel:
•
•
•
•
Are there panel recommendations for additional analyses before presenting these
results to the public?
What key messages are most important for informing the community about these
findings?
Is presenting the percent change in Geometric Mean over time or mean change in
individual PFC level over time best for communicating with the public?
Is the inclusion of Washington County residents in the NHANES subsample concerning?
Should MN Biomonitoring pursue additional information on the Washington County
subsample from NHANES?
5
East Metro PFC3 Biomonitoring Project Results Analysis
Levels of commonly detected PFCs continue to decline in the Original Cohort
Overall, PFC levels in the Original Cohort have continued to decline, supporting findings from
the PFC2 Project that East Metro interventions to reduce PFC exposures through drinking water
are effective. PFOS, PFOA, PFHxS, and PFNA were the most frequently detected PFCs in 2014
(Table 1). Results were compared to PFC1 (2008) and PFC2 (2010) to demonstrate mean
percent change between studies (Table 2).
Geometric Means for PFOS, PFOA, and PFHxS were comparable in the two communities from
which the Original Cohort population was drawn: Oakdale (municipal water) and Lake
Elmo/Cottage Grove (private well water). Percent decline in PFOA and PFHxS were slightly
higher among Oakdale residents compared to Lake Elmo/Cottage Grove residents (data not
shown).
For 26 participants, levels of PFOS, PFOA, and/or PFHxS increased between 2010 and 2014.
Most of these increases were very small and could be explained by uncertainty in laboratory
measurement or ongoing background exposures. Under the 2010/2015 PFOA Stewardship
Program, the major fluoropolymer and telomere producers committed to elimination of PFOA
products by 2015. From 2008 to 2010, PFOA levels increased in 24 participants. From 2010 to
2014, they increased in only 2 participants.
Table 1. PFC levels in Original Cohort (n=156), 2014
n
detect
%
detect
GM
(µg/L)
Median 75 %ile
(µg/L)
(µg/L)
95 %ile
(µg/L)
99 %ile
(µg/L)
Min
(µg/L)
Max
(µg/L)
PFOS
156
100%
18.58
21
34.5
70
93
1
180
PFOA
156
100%
5.45
5.85
11
26
45
0.24
47
PFHxS
155
99%
5.03
5.9
9.1
27
41
<LOD
140
PFBA
52
33%
*
<LOD
0.16
1.2
5.3
<LOD
6.9
PFPeA
0
0%
-
-
-
-
-
-
-
PFHxA
0
0%
-
-
-
-
-
-
-
PFBS
0
0%
-
-
-
-
-
-
-
PFNA
155
99%
0.69
0.67
0.98
2.2
8
<LOD
11
*Not calculated: proportion of results <LOD too high to provide valid result
6
Table 2. Percent change in PFC levels over time, Original Cohort
Mean individual Mean individual Mean individual
percent change percent change percent change
2008 - 2010
2010 - 2014
2008 - 2014
PFOS
-26%
-21%
-45%
PFOA
-21%
-49%
-59%
PFHxS
-13%
-12%
-34%
Figures 1, 2, & 3. 2014 Distributions of commonly detected PFCs in Orginal Cohort
7
Rates of Elimination can be used to demonstrate declines in PFOS, PFOA, and PFHxS
Based on published half-lives of PFOS, PFOA, and PFHxS, we would expect to see a 31%
decrease in PFOS, 53% decrease in PFOA, and 19% decrease in PFHxS in the Original Cohort
between 2010 and 2014 if there were no background sources of exposure.
Using PFC3 results, percent change for these PFCs was first calculated using 2010 and 2014
Geometric Means. This showed declines of 26% in PFOS, 52% in PFOA, and 20% in PFHxS.
Percent change calculated by averaging change in participant’s individual PFC levels showed
slightly smaller declines (Table 3).
Elimination rates for reducing PFC levels by 50% were also calculated using Geometric Means
and individual-level data. Using Geometric Means, elimination rates were 6.3 years for PFOS,
3.2 years for PFOA, and 8.3 years for PFHxS. Using individual PFC results and intervals between
blood draws, elimination rates were 7.2 years for PFOS, 3.4 years for PFOA, and 8.3 years for
PFHxS. These rates should not be directly compared to published half-lives because all sources
of exposure are not known and controlled.
Table 3. Observed and expected percent change, Original Cohort (n=148*)
Percent change in Geometric Mean individual percent Expected percent change**
Mean 2010 - 2014
change 2010 - 2014
2010 - 2014
PFOS
-26%
-23%
-31%
PFOA
-52%
-49%
-53%
PFHxS
-20%
-20%
-19%
*Sample size includes individuals who participated in both PFC2 and PFC3
**Expected percent change calculated using published half-lives of 5.4 years for PFOS, 2.3 and
3.8 years for PFOA, and 8.5 years for PFHxS.
New Resident PFC levels comparable to those of representative sample of the U.S. population
The New Resident group was analyzed for the first time in PFC3 as a check that public health
measures are working to reduce PFC exposures through Oakdale municipal water. We
hypothesized that blood levels of PFOS, PFOA, and PFHxS in Oakdale residents who moved to
the city after the October 2006 intervention would not be significantly different than U.S.
general population levels. Levels were compared to the NHANES 2011-2012 biomonitoring
subsample. Due to the much larger sample size of NHANES (n= 1904), we compared confidence
intervals in lieu of conducting t-tests for differences in PFC levels. For commonly detected PFCs,
all confidence intervals overlapped indicating no differences between PFC3 New Residents and
the NHANES subsample. One exception was PFNA, for which we saw significantly lower levels in
PFC3 New Residents.
9
Table 4. PFC levels in New Residents (n=156), 2014
n
detect
%
detect
GM
(µg/L)
Median
(µg/L)
75 %ile
(µg/L)
95 %ile
(µg/L)
99 %ile
(µg/L)
Min
(µg/L)
Max
(µg/L)
PFOS
156
100%
7.2
7.4
11
21
27
0.34
30
PFOA
156
100%
1.81
2
2.8
5
7.4
0.17
8.1
PFHxS
155
99%
1.63
1.8
2.8
6.3
10
<LOD
19
PFBA
70
45%
*
<LOD
0.22
0.67
0.98
<LOD
4.4
PFPeA
0
0%
-
-
-
-
-
-
-
PFHxA
0
0%
-
-
-
-
-
-
-
PFBS
0
0%
-
-
-
-
-
-
-
PFNA
154
99%
0.48
0.46
0.7
1.3
2
<LOD
3.7
*Not calculated: proportion of results <LOD too high to provide valid result
Table 5. 2014 Levels of PFCs in New Residents compared to NHANES 2011-2012
GM
CI Lower CI Upper Median 75 %ile 95 %ile Range
% detect
PFOS
PFC3
7.2
6.46
8.03
7.4
11
21
.34 - 30
100%
99.6%
6.31
5.84
6.84
6.53
10.5
21.7
<LOD 235
PFC3
1.81
1.62
2.01
2
2.8
5
.17 - 8.1
100%
NHANES
2.08
1.95
2.22
2.08
3.03
5.68
<LOD 43.0
99.5%
6.3
<LOD 19
99%
5.44
<LOD 47.8
98.4%
1.3
<LOD 3.7
99%
2.0
<LOD 80.8
99.3%
NHANES
PFOA
PFHxS
PFC3
NHANES
1.63
1.28
1.42
1.15
1.87
1.43
1.8
1.27
2.8
2.26
PFNA
PFC3
NHANES
0.48
0.88
0.44
0.80
0.53
0.97
0.46
0.86
0.7
1.3
Washington County residents were included in the 2011-12 NHANES sample
10
A small number of Washington County residents were likely included in the 2011-2012
biomonitoring subsample, used here for PFC3 comparisons to the U.S. population. NHANES
randomly samples 15 primary sampling units (PSUs – counties or small groups of contiguous
counties) and enrolls 5,000 participants per year. Washington County was selected as a PSU in
2011; approximately 333 Washington County residents were included in NHANES. As
biomonitoring is done on a random 1/3 subset of participants every 2-year cycle of NHANES,
approximately 111 Washington County residents were in the 2011-12 subsample. Washington
County residents should comprise about 3-6% of the total subsample.
Levels of PFCs are associated with age and sex
Consistent with published PFC literature and our 2008 and 2010 PFC projects, levels of the most
commonly-detected PFCs were associated with age and sex. Unadjusted Geometric Means for
these predictors are shown in Table 6. Age and sex were included as confounders in final linear
regression models.
Table 6. 2014 Unadjusted Geometric Means (µg/L) in New Residents (n=156)
PFOS
PFOA
PFHxS
Age
<35 (n=45)
6.1
1.7
1.3
35-54 (n=65)
6.9
1.7
1.5
>=55 (n=46)
9.0
2.1
2.2
p-value
0.019
0.154
0.0139
Sex
Men (n=61)
10.6
2.3
2.7
Women (n=95 )
5.6
1.6
1.2
p-value
<.0001
0.0007
<.0001
PFNA
0.5
0.5
0.5
0.3224
0.6
0.4
0.0014
Levels of PFCs do not differ between renters and homeowners
The New Residents group was recruited from Oakdale water billing records and select
Washington County Housing and Redevelopment Authority (HRA) properties. We examined PFC
levels by renter status, comparing renters and homeowners. We also examined PFC levels by
HRA status, comparing HRA renters to all other Oakdale New Residents. Geometric Means for
these comparisons are presented in Table 7; no significant associations were seen between
rental status and PFC levels in final adjusted models.
We also examined PFC levels by income, education, and race/ethnicity. A pattern of slightly
higher PFC levels in higher income groups can be seen, but these associations are not
significant. No associations between PFC levels and education were seen in final models.
Comparisons by race/ethnicity were limited to comparing white, non-Hispanic participants to
all other participants. The other group is comprised of participants who identify as Asian,
Black/African American or African, and Hispanic. White, non-Hispanic participants had slightly
higher levels of PFHxS.
11
Table 7. 2014 Adjusted* Geometric Means (µg/L) in New Residents (n=156)
PFOS
PFOA
Residence length
<3 (n=61)
8.06
1.95
3-<5 (n=50)
8.01
1.94
>=5 (n=45)
6.93
1.72
p-value
0.36
0.58
Daily cups of water
1-4 cups (n=72)
7.88
1.72
5-8 cups (n=61)
8.15
2.05
9+ cups (n=23)
6.22
2.02
p-value
0.16
0.27
Renter
No (n= 133)
7.83
1.89
Yes (n= 23)
7.01
1.84
p-value
0.41
0.86
HRA resident
No (n= 138)
7.88
1.90
Yes (n= 18)
6.41
1.71
p-value
0.16
0.52
Income
<$45,000 (n=35)
6.74
1.68
$45,000-$74,999 (n=52)
7.21
1.76
>=$75,000 (n=68)
8.77
2.13
p-value
0.06
0.14
Education
HS degree or fewer years (n=20)
8.78
1.85
Some college or tech degree (n=47)
7.23
1.81
College graduate or more years (n=89)
7.77
1.91
p-value
0.49
0.94
Race/ethnicity
White, non-Hispanic (n=130)
7.95
1.94
Other (n=24)
6.62
1.60
p-value
0.17
0.19
Blood donor
No (n= 132)
7.92
1.90
Yes (n= 24)
6.57
1.77
p-value
0.15
0.63
3M employee
No (n= 143)
7.64
1.84
Yes (n= 13)
8.46
2.31
p-value
0.55
0.24
*Final models adjusted for age and sex
PFHxS
PFNA
1.69
1.93
1.73
0.62
0.55
0.50
0.43
0.11
1.62
1.88
2.09
0.30
0.54
0.45
0.51
0.23
1.79
1.68
0.71
0.50
0.50
0.93
1.81
1.53
0.40
0.50
0.51
0.88
1.36
1.86
1.95
0.08
0.49
0.44
0.55
0.09
1.56
1.64
1.86
0.58
0.51
0.47
0.49
0.90
1.91
1.13
<0.005
0.50
0.49
0.87
1.79
1.56
0.43
0.49
0.46
0.59
1.72
2.46
0.10
0.50
0.51
0.86
12
Levels of PFCs in New Residents are not associated with length of residence in Oakdale
We investigated whether blood levels in New Residents were associated with length of
residence in Oakdale after October 2006, when operations began at the new municipal water
filtration facility. Although PFC levels in Oakdale city water are below health-based limits, low
levels of PFCs are still present in some water samples. This analysis was a check that these low
levels of some PFCs in municipal water are not associated with PFC accumulation in the body.
Spearman correlations between length of residence and PFOS (r=-.03 ), PFOA (r=-.07 ), and
PFHxS (r= .08) were very weak. In the final linear regression models, continuous length of
residence was not associated with log-transformed PFOS, PFOA, or PFHxS after controlling for
age and sex. A weak negative association (β=-.0048, p=.01) was seen between continuous
length of residence and log PFNA. Geometric means by residence group are presented in Table
7. In a logistic regression model, no association was seen between length of residence and
PFBA detection (yes/no).
Figure 4. Scatterplot of PFOS vs. Years in Oakdale
Figure 5. Scatterplot of PFOA vs. Years in Oakdale
No strong associations between PFC levels and blood donation or carpet installation
The PFC2 Project found that Original Cohort participants who donated blood frequently had
lower levels of PFCs compared to people who did not donate blood or did so less often. It also
found that people who had carpet installed in their home in the last year had higher levels of
PFCs. The sample sizes for these groups were small, and PFC levels were still dominated by past
drinking water exposures in the Original Cohort. As length of residence and water consumption
are not associated with PFC levels in New Residents, looking at smaller background exposures
like new carpet should be clearer.
New Residents were not frequent blood donors; only 24 reported donating blood in the last
two years. In final adjusted models, no associations were seen between PFC levels and
dichotomous blood donation (donor/non-donor). Additionally, no associations were seen when
participants were grouped into frequent donors (two or more times per year) and infrequent or
non-donors (less than two times per year).
New Residents were asked about carpet in their office and at home. Among the 94 New
Residents who work in an office, having new carpet installed in the last year was positively
associated with log-transformed PFOA (β=0.4017, p=0.0044). In the larger New Resident group
(n=156), no associations were seen between PFC levels and having carpet installed or cleaned in
the home in the last two years. Possible associations were seen between having carpet treated
at home and log PFOA and PFOS, but the number of participants who reported this was too
small to draw conclusions.
14
Section Overview: Biomonitoring Updates
Biomonitoring Updates are provided in written form. Panel members are invited to ask
questions of staff and comment on all updates.
Information Item:
•
•
MN FEET – Current Status
East Metro Cancer Report – Additional analyses
15
Biomonitoring Updates
MN FEET Current Status
We are currently a few weeks away from the soft launch of the pilot phase of MN FEET, slated
for early June. To that end, active outreach to communities and providers is wrapping up,
translated project materials are printed and interviewer training is scheduled to take place
during the last week of May. Our research partners at HealthPartners have been instrumental
in developing the training, both for the interviewers and in setting up the database for our
project management needs using the secure, web-based system, REDCap. We met with
community leaders in the various groups that we will be recruiting from and will, over the next
few months, follow-up on their suggestions to reach out to community members through
media and in-person opportunities to learn about MN FEET.
After a successful series of meetings to finalize the hospital procedure at Regions Hospital and a
promising meeting with Abbott Northwestern Hospital, we are hopeful that we will have both
hospitals serving as sample collection sites by the end of the pilot phase.
East Metro Cancer Report
At the February, 2015 Advisory Panel Meeting, MDH epidemiologist Kenneth Adams presented
the findings of a new report on cancer occurrence in Washington and Dakota counties and East
Metro communities affected by the PFC drinking water contamination. The report was an
update to a similar analysis done in 2007. The report is now available at the MDH website
2015 Cancer Report
Advisory Panel members recommended additional analyses to help understand the finding of
breast cancer incidence to be higher than expected in the two counties. The following summary
of additional analyses to address the panel’s recommendations is provided below by Kenneth
Adams, PhD, Minnesota Cancer Surveillance System.
Breast cancer incidence in Washington County, Dakota County, and State of Minnesota
On the whole, MCSS analysis of cancer incidence in Washington and Dakota Counties did not
find cancer occurrence to be unusual. Notably however, breast cancer incidence was higherthan-expected in both counties. These results are described in the MCSS report, Data Update:
Cancer Incidence in Dakota and Washington Counties, May 13, 2015.
About 10% more incident breast cancers occurred than expected over the 2003-2012 time
period. The comparison, or reference population for “expected” cancers was the State of
Minnesota. Related findings were that Washington County had the 8th highest female breast
cancer incidence rate among Minnesota’s 87 counties, and Dakota County had the 15th highest
rate.
Discussing the Washington County findings during the meeting, Advisory Panel members
suggested that the higher than expected breast cancer occurrence could be due to high
prevalence of breast cancer screening; i.e., the proportion of females screened for breast
cancer may be higher in Washington County than in other Minnesota counties. The logic is that
breast cancer screening detects early-stage cancer, which is beneficial in preventing breast
16
cancer mortality (this is the intended effect of screening), but this also results in higher
incidence rates.
In response, MCSS staff performed an evaluation breast cancer incidence according to stage at
diagnosis, for Washington County, Dakota County, and the state of Minnesota. The analysis was
descriptive; that is, staff did not evaluate whether the differences found were statistically
significant.
In contrast to the main MCSS report, this analysis included non-invasive, in situ cancer (stage 0).
Breast cancer screening detects tumors that are too small to feel or cause other symptoms and
also abnormal cells lining the breast duct (non-invasive in situ cancer). Screening mammogram
detects these cancers. Cancers detected once they are large enough to be felt or cause other
symptoms are not considered to be screening detected. Localized cancers (stage 1) are
sometimes detected by screening mammogram and other times detected due to signs or
symptoms.
Staff assumed without validation that breast cancer screening in Washington and Dakota
Counties is more frequent than for the state as a whole. Staff also assumed that most stage 0
breast cancer would have been detected by screening, and that stage 1 cancers could be
detected either by screening or diagnosed as the result of signs or symptoms. As a
consequence, interpretation of stage 1 cancers is not straight-forward: they may reflect either
early detection by screening mammography, but may also reflect early diagnosis separate from
screening.
The suggestion that breast cancer screening accounts for higher county-level breast cancer
rates would be supported by data showing higher rates of in situ (stage 0) breast cancer
without higher rates of later stage cancer (stages 2-4 and unstaged cancer). Interpretation of
rates for stage 1 cancer is more challenging, because these rates may or may not reflect the
effects of screening.
Breast cancer incidence appeared higher in Washington and Dakota Counties than in the state
of Minnesota overall, whether considering invasive and non-invasive breast cancer combined,
or invasive breast cancer only (Table 1).
In Washington County the incidence of non-invasive, in situ breast cancer did not appeared to
be similar as in situ breast cancer rates for the state as a whole. This finding contradicts the idea
that breast cancer rates are higher in Washington County than for the state as a whole. On the
other hand, rates of localized breast cancer were higher in Washington County than for the
state as a whole. This result is difficult to interpret since localized breast cancer could have
been detected by either screening or the presence of signs or symptoms. Rates of regional
breast cancers appeared to be higher in Washington County than for the state as a whole.
Overall these results don’t support the idea that higher breast cancer rates in Washington
County are attributable to higher frequency of screening in the county. However interpretation
is complicated by the fact that we don’t know which cases were diagnosed by screening and
which were diagnosed due to signs or symptoms.
17
In Dakota County the incidence of non-invasive, in situ breast cancer did appear higher than in
the state as a whole. The incidence rate of invasive cancer (excluding in situ cancer) also
appeared higher in Dakota County than in the state as a whole, due mostly to higher incidence
of localized invasive cancer.
If in fact breast cancer screening is more frequent in Dakota County than Minnesota as a whole,
the results for in situ cancer provide some support for the idea breast cancer screening
accounts for some of the elevated breast cancer incidence in Dakota County.
18
Table 1. Female breast cancer incidence rates by SEER Summary Stage (as ss_glom), 2003-20121
SEER Summary Stage description
Incidence per 100,000
annually
females
Washington
County
Dakota
County
State of
Minnesota
In situ; noninvasive, intraepithelial
29.8
40.6
30.8
Localized only; confined to breast
87.7
86.0
80.5
Regional
43.1
40.3
38.6
Distant sites or lymph nodes involved
5.5
6.9
6.0
Unstaged
3.8
2.2
3.0
Blank field- No codes assigned in MCSS
0.0
0.0
0.0
117.5
126.6
111.3
52.4
49.4
47.6
Invasive cancers combined (localized and higher)
140.1
135.4
128.1
All stage categories combined
169.9
176.0
158.9
Early stage cancer (in situ and localized)
Late- and unstaged cancer (regional, distant, unstaged)
1Rates
are age-adjusted to female US population at the 2000 census.Results from the main
report were based on invasive breast cancer only. As shown in the table, rates of invasive
cancer were higher in Washington and Dakota Counties, and appreciable proportions of these
rates represent cancer diagnosed at the local stage. It’s not clear the extent to which those
breast cancers diagnosed at the local stage were detected by screening. Consequently, it’s
difficult to answer the question of the extent to which screening accounts for the elevation in
rates of invasive breast cancer in the two counties. In Dakota County, the elevated rates of in
situ cancer offer some support to the idea that rates of invasive breast cancer are –to some
extent- attributable to screening. In Washington County, rates of in situ breast cancer are
similar as in the state as a whole. This suggests that the elevated rates of invasive breast cancer
in Washington County are not the result of higher frequency of screening in the
county.Summary
Task Force members speculated that the higher incidence of invasive breast cancer in
Washington County could be due to higher frequency of breast cancer screening. Higher
frequency of screening could lead to more frequent detection of earlier stage cancers. MCSS
staff tabulated the frequency of breast cancer according to stage at diagnosis.
Rates of in situ, non-invasive breast cancer, which would likely be detected only by screening,
were similar in Washington County as they were for the state as a whole. In contrast, rates of in
situ breast cancer were higher in Dakota County than they were for the state as a whole. These
results do not support the idea that the elevated rates of invasive breast cancer observed in
Washington County are attributable to breast cancer screening. In contrast, the results offer
some support to the idea that some portion of elevation in invasive breast cancer diagnosed in
Dakota County residents is attributable to screening.
Breast cancer mortality in Washington County, Dakota County, and State of Minnesota
MCSS staff calculated breast cancer mortality rates for the years 2003-2012. The analysis was
descriptive; staff did not evaluate statistical significance of results. Task Force members had
suggested that this analysis would show that breast cancer mortality rates in Washington and
Dakota Counties would not be elevated (in contrast to breast cancer incidence rates, which
were elevated). Their logic was that because breast cancer screening prevalence in these
counties is higher than in other Minnesota Counties (which has not been verified by MCSS
staff), and breast cancer screening improves survival and reduces breast cancer mortality, the
higher breast cancer higher incidence in Washington County would not translate to higher
breast cancer mortality.
Results are shown in Table 2. Descriptively, breast cancer mortality in both Washington and
Dakota Counties appeared similar to mortality rates for the State of Minnesota overall.
Table 2. Female breast cancer mortality rates, 2003-2012
Location
Breast cancer mortality
rate*
Washington County
21.2
Dakota County
20.8
State of Minnesota
20.9
*Mortality rates are expressed as deaths per 100,000 females annually,
and are age-adjusted to the 2000 US population.
These results suggest the possibility that breast cancer survival was better in Washington and
Dakota Counties than in the state as a whole. But whether this in turn supports the idea that
the higher breast cancer incidence in Washington County is explained by higher screening
uptake is debatable, since the incidence by stage analyses (Table 1) don’t offer any support for
that theory.
Adams, KA; May 26, 2015
20
Section Overview: State Air and Health Initiative
Jeannette Sample will present highlights from a new technical report soon to be published on
the impacts of air pollution (particulate matter and ozone) on the health of Twin Cities area
residents. Funded by the State Environmental Risks Initiative, the report was a joint project of
the Minnesota Pollution Control Agency air pollution scientists and economists, and
epidemiologists and statisticians from the MN Tracking program. It was modeled after a similar
publication in New York City, and it uses an EPA tool called BenMap (Benefits Mapping) for
quantifying health impacts at the zipcode level.
Chuck Stroebel will give a preview of the “Be Air Aware” website, which was designed by MPCA
and MDH staff and communications specialists to provide useful information for the public,
policy makers and communities.
Questions for the panel:
•
•
•
What key findings from the Public Health Impacts of Air Pollution technical report are
most important for public communications?
How might the results/products of this initiative be used by agencies, organizations, and
our partners?
What suggestions do you have to inform future work on this initiative, including
additional technical analyses and data updates, and/or communications and outreach
activities?
21
State Air and Health Initiative
The State Air and Health Initiative is a joint MPCA/MDH project funded by the 2013 MN
Legislature through the MPCA Environmental Risks Initiative. The goal of this Initiative is to
inform decisions about the health impacts of air pollution, with 3 primary deliverables by July
2015:
Technical Report: “Our Air, Our Health: Public Health Impacts of Air Pollution in the Twin
Cities”, a technical report, prepared by the MPCA and Minnesota Tracking Program at MDH,
provides estimates of the toll of air pollution on health in the 7-county Twin Cities metro area.
The goal of the report is to inform decisions on reducing disease that can be attributable to air
pollution for residents of the Twin Cities metro area and the state of Minnesota.
Breathing polluted air can cause a variety of health problems. While air quality in Minnesota
currently meets federal standards, even low and moderate levels of air pollution can contribute
to serious illnesses and death. The report found that in 2008, 6-12% of all people in the Twin
Cities metro area who died and about 2-5 percent of all people in the Twin Cities metro area
who visited the hospital or emergency room for heart and lung problems did so because fine
particles and ozone made their conditions worse. This roughly translates to about 2,000 deaths
and hundreds of hospitalizations and emergency room visits.
Everyone can be affected by breathing polluted air. This report finds little difference in average
air pollution levels between ZIP codes. Differences between subgroups of the Twin Cities
population in health outcomes related to air pollution exposure largely reflect underlying
vulnerabilities related to age, chronic disease rates, ethnicity, income and other factors.
Improving air quality can provide significant public health benefits. If we reduce fine particles
and ground-level ozone by 10 percent from 2008 levels, we can prevent hundreds of deaths,
hospitalizations, and emergency department visits due to heart and lung conditions every year
that are attributed to these pollutants.
This report provides a local analysis that may be used by state and local decision-makers to
inform actions related to the health impacts of air pollution in the Twin Cities.
Community Toolkit (Be Air Aware web site): Also as a part of this Initiative, MDH and MPCA
are developing a new joint web site with integrated information about air quality and health for
the public. Target audiences for this web site include: individuals and families, businesses and
employers, and local officials and communities. This web site focuses on actions, data and tools
that may be used to protect health and improve air quality in Minnesota. The web content will
include information about indoor and outdoor air quality, and air and health data to inform
decision-making about health disparities and vulnerable populations.
Health Impact Assessment: Also through this Initiative, the MDH Environmental Health Division
is conducting a Health Impact Assessment in South Minneapolis. The HIA, driven by stakeholder
involvement from a variety of community-based organizations, will focus the potential health
impact of the Green Zone Initiative, an initiative of the City of Minneapolis. The Green Zone
Initiative provides a process for green zone designations for neighborhoods or clusters of
neighborhoods that face the cumulative impacts of environmental, social, political and
22
economic vulnerability. The final results and recommendations of the HIA will be shared with
stakeholders and local officials to inform decisions regarding health impacts of implementation
of green zone designations in the city.
Next Steps: MDH and MPCA have developed a coordinated communications plan for the State
Air and Health Initiative. The technical report and community toolkit (web site) will be released
together in early July 2015. Additional funding from the 2015 MN Legislature likely will be used
to gather input from stakeholders and communities, and to update and enhance the products
of this Initiative.
Also related to the Initiative, Minnesota Tracking recently updated maps for asthma and
chronic obstructive pulmonary disease (including ZIP code hospitalization/ED data for the 7county metro area). These maps, as well as data on population characteristics (poverty, income,
access to health care) are available through the MN Public Health Data Access portal:
•
•
Asthma (interactive maps - Asthma) (static maps - Asthma)
Chronic Obstructive Pulmonary Disease (interactive maps - COPD)
23
Section Overview: Tracking Updates and Program Evaluation
Matthew Montesano will demonstrate the new data visualizer and describe the MN Tracking
program evaluation. Portal updates on heat-related illness maps, folic acid use data, birth
defects data, air, health and poverty data visualizer tool, asthma 2103 emergency department
and hospitalizations data, and youth exposure to environmental tobacco smoke, smoking and
obesity are provided in written form. Panel members are invited to ask questions and comment
on all updates.
Information item updates:
•
•
•
Portal Updates
Strategic Planning Meeting
Upcoming Grant Year Plans
24
Tracking Updates and Program Evaluation
The following topics were updated on the Portal:
•
Launched interactive maps of heat-related illness, and maps of populations vulnerable
to heat-related illness.
• Launched Folic Acid Use data, part of the Birth Defects topic, along with an update to
birth defects data.
• Soft-launched the air, health, and poverty data visualizer tool.
• Updated Asthma with 2013 ED and hospitalizations data, a new and updated data
query, and a new interactive mapping system.
• Updated youth exposure to environmental tobacco smoke; smoking; obesity.
The tracking team is currently working on adding new years of data to cancer, COPD, heart
attacks, drinking water quality, arsenic in private wells, and reproductive-and birth-outcomes.
Strategic Planning Meeting
In May 2015 the CDC National Tracking Network sponsored a workshop for grantees (principal
investigators, program managers) and CDC to discuss strategic planning for 2016-2020. This
workshop provided a unique opportunity to discuss accomplishments and challenges for the
Tracking Network and planning to address emerging issues. The following are a few highlights:
John Hopkins staff provided an overview of an Expert Panel convened in March 2015 to gather
input on the challenges and accomplishments of the Tracking Network. Participants in this
Panel included some grantees (invited), CDC, and John Hopkins staff had played a central role in
making the case for establishing the Tracking Network over a decade ago. The Panel identified
several accomplishments and challenges for the Tracking Network. Recommendations from the
Panel, as well as additional input from external partners/organizations, will be summarized at
the next Tracking Network Grantee Meeting in Atlanta, GA in August 2015.
Strategic planning sessions also included roundtable discussions that identified challenges and
opportunities for the Tracking Network. Participants expressed the desire for more flexibility to
determine requirements for state tracking data portals; concern about duplication of efforts
and the work necessary to maintain and update multiple data/measures for CDC national and
state portals; and interest in focusing more on environmental epidemiology capacity and data
analysis/utilization. Several participants also expressed concern about finding a balance
between offering breadth vs depth in terms of the data and expertise provided by the Tracking
Network. Participants highlighted the importance of communicating the impact of Tracking
through success stories (or public health actions) to CDC management, and congressional
delegation.
A summary of the strategic planning sessions and results of the Expert Panel will be reported at
the upcoming CDC National Tracking Network grantee meeting in August 2015.
Plans for the MN Tracking Program, Upcoming Grant Year
The Minnesota Tracking Program submitted the work plan and budget to CDC for the next grant
year, starting August 1, 2015. Items in this work plan include:
25
Content areas: maintain and update content on the MN Public Health Data Access portal;
develop and explore new content at fine spatial resolution: childhood obesity and implement
public water systems; cancer and explore birth outcomes; explore and evaluate new content:
traffic, radon, youth smoking; update childhood lead poisoning data and measures.
Information Technology actions: maintain and update content on the Data Access portal in the
form of charts, maps, queries and profiles; implement the work plan for new content, as
developed by Tracking and technical upgrades.
Data utilization tasks: update data and analysis for the State Air and Health Initiative; promote
data utilization in MN Health Impact Assessments and explore innovative ways of visualizing
data, such as co-displays of data similar to the new air, health, and poverty data visualizer.
Communications actions include: communicate and share stories about the impact of tracking;
how data are being used to inform public health action; maintain outreach to target audiences,
including local health departments through portal demonstrations, webinars, presentations,
and bulletin updates to email subscribers; expand outreach to new target audiences, including
health care organizations and accountable care.
Collaboration efforts include: continue contract and project with the Great Lakes Inter-Tribal
Epidemiology Center and the Fond du Lac Tribe in MN; implement the Rapid Response Roster, a
tool for handling long-term issues follow disasters; continue collaboration with MPCA,
Environmental Risks Initiative; continue collaboration with MDH Environmental Health Division,
for example, climate change, Health Impact Assessment, new content e.g., radon.
Evaluation components include: tracking and report performance measures, based on web and
subscriber analytics, key informant interviews, and usability testing.
26
Section Overview: Mercury Impact Analysis for Informing Reduction
Initiatives
Jean Johnson will introduce a recent MN Tracking analysis of the health and economic burden
of mercury in newborns. Blair Sevcik will describe the methods used.
This analysis is a supplement to the current report now available on the MN Tracking website.
“The Economic Burden of the Environment on Two Childhood Diseases: Asthma and Lead
Poisoning in Minnesota”
Similar to the previous report, this analysis documents the economic cost of elevated mercury
in newborns born in one year using national biomonitoring data from 2011-2012. It calculates
the impact on IQ loss over a lifetime, and it estimates the fraction that is attributed to
anthropogenic sources of mercury in the environment, which can be addressed through
environmental interventions or policies.
The analysis is intended to inform decisions by the public, policy-makers and advocacy groups
so that policy and resources are directed towards actions that will both reduce the health
impact and save money.
Frank Kohlasch, Section Manager of the Environmental Analysis and Outcomes Section at the
Minnesota Pollution Control Agency, will describe the Statewide Mercury Reduction Initiative,
important to reducing these impacts and how this analysis may inform their work.
This section includes two background documents:
1. A portion of the draft report “The Economic Burden of the Environment on Children’s
Health: The Cost of Prenatal Mercury Exposure” that describes methods and limitations
2. MPCA: Actions Addressing Mercury in Minnesota’s Environment
Questions for the panel:
•
•
•
How can MDH-EHTB continue to support and inform the MPCA’s mercury reduction
initiatives?
Given the limitations, how well does the economic burden analysis serve its intended
purpose?
What additional information would be most helpful going forward?
Mercury Impact Analysis for Informing Reduction Initiatives
THE FOLLOWING REPORT IS A PRELIMINARY DRAFT FOR
REVIEW ONLY. PLEASE DO NOT COPY OR DISTRIBUTE.
The Economic Burden of the Environment on Children’s Health:
The Cost of Prenatal Mercury Exposure
May 20, 2015
Introduction
This report estimates the health impact and economic cost of prenatal mercury exposure
attributable to human-caused sources of mercury in the environment. The Minnesota
Biomonitoring Program is studying the amount of mercury exposure in newborns, and potential
exposure disparities, in Minnesota communities.
Babies in utero are exposed to mercury through their mothers’ blood, which passes through the
placenta. Developing babies are most at risk because small amounts of mercury can damage
the developing brain and nervous system. Mercury exposure can affect a child's learning
abilities, memory, and attention, and lead to learning problems later in life. In adults, ongoing
exposure to mercury can damage the kidneys and nervous system.
For newborns, maternal consumption of fish that contain mercury is thought to be the major
source of prenatal exposure to mercury. Older, larger fish and fish that eat other fish have the
highest levels of mercury. Methylmercury is an organic mercury compound that is produced by
bacteria from other forms of mercury in rivers, lakes and oceans, and is the most toxic form of
mercury.
Pregnant women can also be exposed to inorganic or elemental mercury from broken
thermometers in the home, skin-lightening creams that contain mercury, dental fillings and
mercury used for ritual and folk purposes. Although more study is needed, inorganic mercury is
known to cross the placenta and has been found in cord blood (U.S. Environmental Protection
Agency, 2007).
The source of nearly all mercury in Minnesota waters (99%) is deposition from the atmosphere.
Anthropogenic sources of mercury include emissions from energy production, material
processing (mostly taconite), industrial boilers, refineries, recyclers, solid waste processors,
smelters, crematories, product manufacturing, incinerators, and use of mercury in products
(Minnesota Pollution Control Agency, 2009). Ninety percent of human-caused mercury
deposition in Minnesota comes from sources outside the state.
This report builds upon a previously published report of the MN Environmental Public Health
Tracking (MN Tracking) program (Minnesota Department of Health, 2014), which estimated the
economic burden of two important environmentally-related health conditions in children:
asthma and blood lead poisoning. The purpose of these reports is to use state and national
biomonitoring data (measuring chemicals in people) and MN Tracking data in ways that help
inform policy-makers about the health and economic benefits of policy, actions and
interventions for protecting the health of future generations. Policy initiatives will be described
in a future section, titled “Actions Addressing Prenatal Mercury Exposure in Minnesota.”
Background
Mercury exposure trends in the U.S.
The proportion of women of childbearing age with elevated mercury levels has declined
Percent above 5.8µg/L
8
7
6
5
4
3
2
1
0
1999-2000 2001-2002 2003-2004 2005-2006 2007-2008 2009-2010
Data are for women age 16-49 and are from a paper by Birch
et al. (2014) that used two-year cycles of NHANES data.
Only limited data exist on mercury levels in newborns, but we can examine trends in women of
childbearing age from the National Health and Nutrition Examination Survey (NHANES), a
nation-wide survey that includes biomonitoring. The EPA’s reference level for mercury is the
level beyond which it is thought to have harmful effects on health. The percent of women aged
16-49 with total blood mercury levels higher than the EPA’s reference level for methylmercury
(5.8 micrograms of mercury per liter of blood or µg/L) declined from 1999-2000 to 2003-2004
and has since remained relatively stable. Just over 2% of U.S. women of childbearing age had
blood mercury levels that exceed the reference dose in 2009-2010.
29
Disparities observed
The proportion of women of childbearing age with elevated mercury levels differs by race
Percent above 5.8µg/L
All women = 3.4%
14
12
10
8
6
4
2
0
Other race
Black, non- White, nonHispanic
Hispanic
Other
Hispanic
Mexican
American
Data are for women age 16-49 and are from a paper by
Birch et al. (2014) and uses 1999-2010 NHANES data.
Exposure to mercury differs by race/ethnicity. Most strikingly, 12.0% of women who identified
themselves as “other race” (includes Asian, Native American, Pacific Islander, and Caribbean)
exceeded the EPA reference dose level (above 5.8 µg/L).
Similar percentages of black, white, and Hispanic women exceeded this level compared to the
average for all women (3.4%). A lower percentage (0.7%) of Mexican American women
exceeded this level. It’s possible that these race/ethnicity disparities are related to fish
consumption patterns.
Studies of cord blood in newborns in the U.S. have found marked disparities by race/ethnicity in
exposure to mercury. Certain groups, such as some Asian populations and African Americans,
have been shown to have higher exposures than white newborns (King et al., 2013; Lederman
et al., 2008). Although not a measure of newborn exposure, biomonitoring studies have also
found elevated urinary mercury levels in Latina women who used skin-lightening creams
(McKelvey, Jeffery, Clark, Kass, & Parsons, 2011; Weldon et al., 2000).
Information on newborn mercury exposures in Minnesota is limited. An MDH study of cord
blood collected from babies born to primarily white, affluent, urban mothers in the Twin Cities
found 2% of participants had exposures above a 5.8 µg/L. A study in the Lake Superior region of
the state tested 1,100 newborn blood spots for total mercury and found that 10% of those
tested had levels above 5.8 µg/L (Minnesota Department of Health, 2011). Current
biomonitoring studies are expected to provide additional estimates.
The Costs of Prenatal Mercury Exposure: Methods & Results
This report adopts methods described in Trasande & Liu (2011) and Landrigan et al. (2002) and
applies these methods using current Minnesota population statistics, U.S. population
biomonitoring data, and economic information.
30
The formula
The formula for estimating the economic burden of environmentally-related disease described
by Landrigan, et al. (2002) is:
Economic burden = case counts x environmentally attributable fraction (EAF) x cost per case
Economic burden is estimated as the number of cases of disease in a defined population and
time period, multiplied by the environmentally attributable fraction (EAF) and the cost per
disease case.
Estimating elevated mercury exposure cases
Research (Stern & Smith, 2003) has shown that the average ratio of mercury levels in a
newborn’s cord blood compared to maternal blood is 1.7. Therefore, Trasande et al. (2011)
applied this 1.7 ratio to the U.S. EPA’s reference dose of 5.8 µg/L and arrived at a level of 3.4
µg/L in women of childbearing age as a threshold level for estimating the number of elevated
newborn mercury exposure cases.
From the most recent 2011-2012 cycle, NHANES measured the level of mercury in blood from a
random sample of about 1,600 U.S. women of childbearing age (16-49 years). About 8.6% of
women of childbearing age sampled nationwide had a total mercury level above the threshold
of 3.4 µg/L (138 women out of about 1,600 women overall). We applied this percentage to the
total number of births in Minnesota in 2014 to estimate the percent and number of newborns
with elevated exposure. In 2011 and 2012, an average of 35,119 boys and 33,480 girls were
born to Minnesota women per year. Therefore, approximately 3,020 boys and 2,879 girls were
potentially impacted by prenatal mercury exposure each year. (Table 1)
Table 1: The estimated number of Minnesota babies born each year with elevated mercury
levels, 2011-2012.
Proportion of children born to
Average number of children
Estimated number
women with mercury levels
born per year in 2011-2012
of children affected
above threshold
8.6%
Boys
35,119
3,020
8.6%
Girls
33,480
2,879
IQ point loss attributed to blood mercury levels
Among women with an elevated mercury level in NHANES, the average mercury concentration
in blood was 5.27 µg/L (95% confidence interval: 4.69-5.92 µg/L). We used this level for
calculating average exposure and IQ points lost among Minnesota newborns as described
below.
Axelrad et al. (2007) determined that 0.18 IQ points are lost per every unit increase (1 µg/L)
mercury in a newborn mother’s blood above a threshold of 3.4 µg/L. That means if a woman
has a mercury level of 9 µg/L (5.6 units above the threshold), her developing baby’s IQ level
would theoretically decrease by a full IQ point (5.6 x 0.18).
31
Given our estimate that the average mercury level in the blood of newborns born with
exposures above the threshold is 5.27 µg/L, and that this represents an average unit increase of
1.87 µg/L above 3.4, we calculate an average IQ loss of 0.34 points due to mercury exposure (all
sources) for babies in Minnesota with elevated exposure (8.6% of births.)
The environmentally attributable fraction (EAF)
The environmentally attributable fraction (EAF) estimates the fraction of the disease that would
be avoided or eliminated if the modifiable environmental risks were removed or reduced to the
lowest level possible. Based on the 1997 Mercury Study Report to Congress (U.S. EPA, 1997),
Trasande et al. (2011) determined that 70% of mercury exposure in people can be attributed
anthropogenic sources. This estimate is consistent with a 1992 report of mercury accumulation
in lakes in the upper Midwest of the US which found that natural atmospheric mercury
concentrations were only about 25% of modern levels (Swain, Engstrom, Brigham, Henning, &
Brezonik, 1992). Global and regional anthropogenic emissions combined are estimated to
account for about 70% of mercury entering Midwestern lakes (Engstrom & Swain, 1997; Swain,
Engstrom, Brigham, Henning, & Brezonik, 1992). Accordingly, this report uses an EAF of 70%.
Given our estimate that the average IQ loss for babies born with elevated exposure is 0.34
points, and 70% can be attributed to human-caused sources, we estimate that the average IQ
loss attributed to anthropogenic sources is .70 x 0.34, or 0.24 points lost. (Table 2).
Table 2: IQ points lost per child due to elevated mercury levels, 2011-2012.
Average
Units of
IQ points lost
Average IQ points EAF
Environmentallymercury level
mercury
per unit
lost for children
attributable IQ points lost
above
above 3.4 increase above
above mercury
for children above
threshold
µg/L
3.4 µg/L
threshold
mercury threshold
5.27 µg/L
(range:
4.69-5.92 µg/L)
1.87 µg/L
0.18
0.34
70%
0.24
Estimating lifetime earnings capacity and loss
Landrigan et al. (2002) estimates 2.39% of potential lifetime earnings are lost for every
decrease in an IQ point. Given 0.24 IQ points are lost on average, this equates to an estimated
0.57% of lifetime earnings lost, on average, for babies born in 2011-2012 in Minnesota with
elevated mercury exposure at birth.
Lifetime earning capacity was provided from market productivity estimates for boys and girls
published in Grosse et al. (2009), but did not include household productivity. We chose lifetime
earnings with 1% growth and 3% discount rate for children aged 0-4 years: $1,203,318 for boys
and $709,824 for girls in 2014 dollars (Grosse, et al., 2009). The original estimates were for
2007 dollars and we adjusted to 2014 dollars using the Consumer Price Index calculator (Bureau
of Labor Statistics).
We find that human-caused prenatal mercury exposure in Minnesota cost $20,846,354 for all
boys and $11,723,141 for all girls born within a single year in Minnesota due to lost lifetime
32
earnings. Therefore, the total economic burden of environmentally attributable mercury
exposure in utero to Minnesota children born in 2011 or 2012 adds up to $32,569,495 in 2014
dollars of lost lifetime earnings for each year of babies born. Given the 95% confidence interval
for the average mercury level above the threshold in Table 2, the total economic burden ranges
from $22.1 million to 43.1 million. (Table 3)
Table 3: Total lifetime earnings lost attributed to elevated mercury levels for children born in
Minnesota, 2011-2012.
Lifetime
Lifetime
Lifetime
0.57% of Estimated
Lifetime
Total
earnings
earnings
earnings
lifetime number of earnings lost
economic
lost for
lost for
per child
earnings children
burden
every IQ
0.24 IQ
per child affected
(2014$)
(from Table 1)
point
points
2.39%
0.56%
Boys
$1,203,318
2.39%
0.56%
Girls
$709,824
$6,902
$4,072
3,020
$20,846,354
$32,569,495
(range: $22.143.1 million)
2,879
$11,723,141
$32,569,495
(range: $22.143.1 million)
An alternative method for estimating the cost of prenatal mercury exposure would calculate an
IQ deficit attributable to only methyl mercury levels from NHANES, rather than total mercury
levels. The proportion of women with methyl mercury levels above the threshold decreases to
6.3%. Taking the average methyl mercury level among these women, the total economic
burden was lower ($22.9 million in 2014 dollars).
Limitations of this analysis
Costs likely underestimate the true economic burden
This report only measures the cost of lifetime earnings lost due to mercury exposure’s impact
on IQ. It does not include other costs that are difficult to measure, such as decreased quality of
life, increased crime, or increased use of education services. Lifetime earnings in this report are
limited to market productivity, and do not include the cost of mercury’s impact on household
productivity. If both market and household productivity were included in the cost, the impact of
total mercury on lifetime earnings would amount to $40.7 million (about $8 million higher).
Full scale IQ is recognized as a composite index of cognitive function, predictive of later
academic and occupational success (Neisser et al., 1996). However, it does not include all
neurodevelopmental deficits associated with mercury, such as effects on motor skills and
attention/behavior. Furthermore, the adverse impact of maternal mercury concentration on
full scale IQ may be underestimated because concurrent consumption of fish fatty acids might
enhance cognitive development (Pichery et al., 2012; Rice, Hammitt, & Evans, 2010). If
exposure is assumed to be from fish consumption then the benefits of fatty acids found in fish
need to be included in assessment of changes in IQ (FAO/WHO, 2011).
33
The environmentally attributable fraction (EAF) is an uncertain estimate
The EAF for mercury exposure (70%) used in this report is based on a report by Trasande et al.
(2011) and recent studies estimating the contribution of anthropogenic sources to
environmental mercury in the Midwestern U.S. (Engstrom & Swain, 1997; Swain, et al., 1992).
The level of mercury exposure in Minnesota newborns is unknown
Blood mercury levels in U.S. women of childbearing age from NHANES are used in this report to
estimate prenatal exposure in Minnesota newborns. Research (Mahaffey, Clickner, & Jeffries,
2009) using national data show that blood mercury levels are, on average, the lowest among
Midwest women, compared to women from other U.S. regions but also that women in the
Great Lakes Coast region may have higher blood mercury levels than nationally. Findings from a
recent study in Minnesota (Minnesota Department of Health, 2011) also suggest that some
populations in Minnesota may have higher levels and that exposure may vary seasonally.
More study is needed in Minnesota to know whether the published national measures of the
proportion of women of childbearing age with mercury levels above a threshold are an accurate
representation of Minnesota women. A direct measure of levels in Minnesota newborns (e.g.
cord blood) would provide a better estimate for determining the impact of prenatal mercury
exposure in the state.
The burden is not shared equally
Some Minnesota communities or people of various racial/ethnic backgrounds may have higher
exposures to mercury, and would share a larger economic burden due to this exposure. These
populations may include communities that have higher fish consumption (especially local,
subsistence fishing) such as Hmong and other Asian women, and those who may use skinlightening creams containing mercury such as Latina and Somali women. No biomonitoring data
are available on mercury exposures in these groups of Minnesotans.
Some Minnesotans may also be more vulnerable to the health and economic effects of mercury
exposure due to inequities in other factors like health care access, housing conditions, or early
childhood education.
Lifetime earnings used in this report are for the U.S. population of children aged 0-4 years.
Lifetime earnings estimates for Minnesota may be different than the nationwide estimate, and
children of different race/ethnicities may have higher or lower lifetime earnings than average.
The next step in this analysis is to repeat the calculations using mercury levels from different
race/ethnicity groups using 2011-2012 NHANES data.
References
Axelrad, D., Bellinger, D., Ryan, L., & Woodruff, T. (2007). Dose-response relationship of
prenatal mercury exposure and IQ: an integrative analysis of epidemiologic data.
Environmental health perspectives, 115(4), 609-615.
Birch, R., Bigler, J., Rogers, J., Zhuang, Y., & Clickner, R. (2014). Trends in blood mercury
concentrations and fish consumption among U.S. women of reproductive age, NHANES,
1999–2010. Environmental research, 133, 431-438. doi: 10.1016/j.envres.2014.02.001
34
Bureau of Labor Statistics. BLS Inflation Calculator. Retrieved March, 2014, from CPI Inflation
Calculator
Engstrom, D., & Swain, E. (1997). Recent Declines in Atmospheric Mercury Deposition in the
Upper Midwest. Environmental Science & Technology, 31(4), 960-967.
FAO/WHO. (2011). Report of the Joint FAO/WHO Expert Consultation on the Risks and Benefits
of Fish Consumption: Food and Agriculture Organization of the United Nations; World
Health Organization.
Grosse, S. D., Krueger, K. V., & Mvundura, M. (2009). Economic productivity by age and sex:
2007 estimates for the United States. Medical care, 47(7 Suppl 1), S94-103. doi:
10.1097/MLR.0b013e31819c9571
King, E., Shih, G., Ratnapradipa, D., Quilliam, D., Morton, J., & Magee, S. (2013). Mercury, lead,
and cadmium in umbilical cord blood. J Environ Health, 75(6), 38-43.
Landrigan, P. J., Schechter, C. B., Lipton, J. M., Fahs, M. C., & Schwartz, J. (2002). Environmental
pollutants and disease in American children: estimates of morbidity, mortality, and costs
for lead poisoning, asthma, cancer, and developmental disabilities. Environmental
health perspectives, 110(7), 721-728.
Lederman, S., Jones, R., Caldwell, K., Rauh, V., Sheets, S., Tang, D., . . . Perera, F. (2008). Relation
between cord blood mercury levels and early child development in a World Trade
Center cohort. 116, 8(1085-1091). doi: 10.1289/ehp.10831
Mahaffey, K., Clickner, R., & Jeffries, R. (2009). Adult Women’s Blood Mercury Concentrations
Vary Regionally in the United States: Association with Patterns of Fish Consumption
(NHANES 1999–2004). Environmental health perspectives, 117(1), 47-53. doi:
10.1289/ehp.11674
McKelvey, W., Jeffery, N., Clark, N., Kass, D., & Parsons, P. (2011). Population-based inorganic
mercury biomonitoring and the identification of skin care products as a source of
exposure in New York City. Environmental health perspectives, 119(2), 203-209. doi:
10.1289/ehp.1002396
Minnesota Department of Health. (2011). Mercury Levels in Blood from Newborns in the Lake
Superior Basin (GLNPO ID 2007-942),
Minnesota Department of Health. (2014). Economic Burden of the Environment.
Minnesota Pollution Control Agency. (2009). Implementation Plan for Minnesota’s Statewide
Mercury Total Maximum Daily Load.
Neisser, U., Boodoo, G., Bouchard, T. J., Boykin, A. W., Brody, N., Ceci, S. J., . . . Urbina, S.
(1996). Intelligence: Knowns and unknowns. American Psychologist, 51(2), 77-101. doi:
10.1037/0003-066X.51.2.77
Pichery, C., Bellanger, M., Zmirou-Navier, D., Fréry, N., Cordier, S., Roue-Legall, A., . . .
Grandjean, P. (2012). Economic evaluation of health consequences of prenatal
methylmercury exposure in France. Environ Health, 11, 53. doi: 10.1186/1476-069X-1153
Rice, G., Hammitt, J., & Evans, J. (2010). A probabilistic characterization of the health benefits of
reducing methyl mercury intake in the United States. Environmental Science &
Technology, 44(13), 5216-5224. doi: 10.1021/es903359u
35
Stern, A., & Smith, A. (2003). An assessment of the cord blood:maternal blood methylmercury
ratio: implications for risk assessment. Environmental health perspectives, 111(22),
1465-1470. doi: 10.1289/ehp.6187
Swain, E., Engstrom, D., Brigham, M., Henning, T., & Brezonik, P. (1992). Increasing Rates of
Atmospheric Mercury Deposition in Midcontinental North America. Science, 257(5071),
784-787.
Trasande, L., & Liu, Y. (2011). Reducing the staggering costs of environmental disease in
children, estimated at $76.6 billion in 2008. Health affairs, 30(5), 863-870. doi:
10.1377/hlthaff.2010.1239
U.S. Environmental Protection Agency. (2007). TEACH Chemical Summaries: Mercury
(Inorganic): U.S. EPA's Toxicity and Exposure Assessment for Children's Health (TEACH).
U.S. EPA. (1997). Mercury Study Report to Congress.
Weldon, M., Smolinski, M. M., A, Hasty, B., Gilliss, D., Boulanger, L., Balluz, L., & Dutton, R.
(2000). Mercury poisoning associated with a Mexican beauty cream. West J Med,
173(1), 15-19.
36
Addressing Mercury in Minnesota’s Environment
Statewide mercury reduction efforts have been underway since 1990 1. For example,
Minnesota’s schools are now mercury free zones, our electric utilities have reduced mercury
emissions by nearly 90% from 1990 levels, and there are bans in place on the sale of mercury
containing devices. As of 2011, Minnesota’s total mercury emissions have decreased by 76
percent compared to the 1990 baseline.
Not only is there a risk to human health from exposure to mercury vapor or methyl mercury via
fish consumption, but the risk to wildlife from consumption of contaminated fish is likely even
greater, largely because wildlife do not alter their food intake in response to consumption
advice . 2 According to 2014 data, 97% of 490 stream and river sections assessed, and 95% of
1,214 lakes assessed are impaired for exceeding safe levels of mercury in fish tissue.
Minnesota’s Pollution Control Agency (MPCA), Department of Natural Resources and
Department of Health collaborate to monitor mercury in fish.
The state of Minnesota has formally adopted a plan to reduce mercury emissions in the state
from 1990 levels by 93% by year 2025. If the plan is fully implemented, including similar
reductions from national and global sources, 90% of our lakes and streams with mercury
impairments will meet clean water standards for mercury levels in fish. The other 10% of those
lakes and streams are more efficient at concentrating (i.e. bioaccumulating) mercury in the
food chain, and therefore will need additional actions (or more time) to fully achieve the
standards. The MPCA is currently researching mercury cycling in several streams to determine
what factors make the bioaccumulation of mercury more efficient in those ecosystems.
Most recently, a new air emissions mercury rule was enacted in 2014 for the state of Minnesota
that requires mercury reduction planning, emissions reporting, and performance standards for
some mercury emission sources. The MPCA continues to build relationships with industry
representatives, environmental groups, and local and national government representatives to
work together to reduce mercury contamination in Minnesota. To learn more, visit MPCA’s
mercury website.
1
2
MPCA. 1994. Strategies for Reducing Mercury in Minnesota. Minnesota Pollution Control Agency. 54 pp.
Swain, E.B., et. al. Socioeconomic Consequences of Mercury Use and Pollution. AMBIO. 36(1):45-61.
Section Overview: Other Information
This section contains documents that may be of interest to panel members.
•
•
•
•
•
2015 Upcoming Advisory Panel Meeting dates
February 10, 2015 Advisory Panel Meeting Summary
Advisory Panel Roster
Biographical Sketches of Advisory Panel Members
Biographical Sketches of Staff
38
2015 Advisory Panel Meetings
Tuesday, October 13, 2015
1-4 pm
All meetings for 2015 will take place at
The American Lung Association of Minnesota
490 Concordia Avenue
St. Paul, Minnesota
38
February 10, 2015 Advisory Panel Meeting Summary:
Environmental Health Tracking & Biomonitoring
1:00–4:00 p.m., American Lung Association
Advisory Panel Members: Bruce Alexander, Fred Anderson, Melanie Ferris, Jill Heins Nesvold,
Patricia McGovern, Geary Olsen, Gregory Pratt, Cathy Villas-Horns, Lisa Yost,
Advisory Panel Regrets: Alan Bender, Thomas Hawkinson and Steven Pedersen
MDH staff: Kenneth Adams, Betsy Edhlund, Carin Huset, Jean Johnson, Jim Kelly, Tess Konen,
MaryJeanne Levitt, Mary Manning, Pat McCann, Matthew Montesano, Paul Moyer, Jessica
Nelson, Christina Rosebush, Lucy Ross, Paul Swedenborg, Janis Taramelli, Addis Teshome, and
Dan Tranter
Welcome and introductions
Chair Pat McGovern welcomed the attendees and invited the panel members and audience to
introduce themselves. Jean Johnson, Director of the Environmental Public Health Tracking and
Biomonitoring Program, informed the panel of David DeGroote’s appointment expiring on the
first of the year. Although David had reapplied for his seat, the appointing authority was the
House of Representatives, and the Secretary of State’s Office had not yet received the House’s
choice for the panel. Jean also notified the panel of two articles published in the December
2014 Journal of Environmental Health. Both articles involved the PFC Biomonitoring project and
appeared in the Advancement of the Science section of the journal. The first article was entitled
Biomonitoring for Perfluorochemicals in a Minnesota Community With Known Drinking Water
Contamination, and the second article, “Communicating About Biomonitoring and the Results
of a Community-Based Project: A Case Study on One State’s Experience.” The journal is
published by the National Environmental Health Association. Pat McGovern led the panel in a
congratulatory round of applause for the work that went into the journal articles.
Blood Spot Project Results Update
Background materials for this presentation were found on pages 5-8 of the February 10, 2015
Advisory Panel book. Jessica Nelson introduced Addis Teshome, epidemiologist with MN
Biomonitoring. Together they presented updates on three MN Biomonitoring projects that
were using available specimens from other studies in Minnesota to investigate mercury levels in
newborns and pregnant women.
Jessica reviewed the reason for doing the studies as a follow up to the Mercury in Lake Superior
Newborns Study. The two main reasons were to assess whether measuring mercury in newborn
bloodspots was a reliable way to estimate newborn exposure to mercury, and to explore the
extent of newborn exposure to metals in Minnesota. They were essentially validating the
question of how did measurements in the bloodspots compare to measurements in other
commonly used biomarkers.
39
The first study, the Pregnancy & Newborns Exposure Study, which was part of the larger
University of Minnesota TIDES study, measured a small urban sample (48) of paired newborn
bloodspots and cord blood. The second study, the Riverside Newborn Mercury Project, also
part of a larger University of Minnesota Riverside Birth Study (RBS), sampled from the same
clinical population as the TIDES study and had new results. Finally, there was the National
Children’s Study, which had three samples from the mom-baby pairs: the newborn bloodspot,
cord blood and maternal blood. When cord blood was tested, it was also tested for lead and
cadmium, as well as mercury, something that was not yet available for bloodspots.
The Riverside Newborn Mercury Study, a University of Minnesota study, recruited pregnant
women receiving prenatal care and giving birth at Fairview Riverside Hospital from 2008 to
2010. The goal of the study was to characterize newborn mercury exposures in various
Minnesota communities. The women filled out questionnaires and provided specimens
specifically for this study. Newborn bloodspot samples (160) were sent to the Minnesota
Department of Health Public Health Laboratory for routine metal analysis using ICP-MS. An
unexpected finding was that mercury was detected in blanks taken from bloodspot filter paper
of 11 samples, or 7%. There was a concern that this was related to a sticker on the filters, but
further investigation revealed that this was not the case. 63 samples or about 40% of the
samples were below the MDL and so they had been assigned a value equal to MDL/√2.
Addis presented a table comparing the results of newborn blood spot testing and explained
that the RBS and TIDES studies were quite similar. Of importance, she noted, was that there
were a lower proportion of samples with levels above the 5.8 reference level (1.3%), with no
drastically high levels, although the RBS had a relatively limited dataset. Addis continued that
the next steps involved analyzing the association with demographic factors and survey
responses and exploring combining the RBS data with TIDES data.
Jessica reminded the panel that the Pregnancy and Newborns Exposure Study found that
newborn bloodspot mercury levels on average were lower than cord blood levels, with a ratio
of 0.85 ±0.4. Split lab experiments revealed that differing lab methods may have accounted for
any discrepancy. The findings were limited by a small sample size, particularly those 16 samples
with metal detected in spot and cord blood.
Jessica asked Chair Pat McGovern to update the panel on the National Children’s Study status,
as she had conducted the Minnesota portion of the study. Pat explained that the study
convened an advisory panel to look at recommendations of the study going forward. They
decided if they could not agree on a design going forward, they would rechannel the funding to
projects involving children’s environmental health, effectively ending the controversial study.
Jessica added that the program office was still working with MDH -Environmental
Epidemiology; they wanted to be sure that the projects underway could be completed, while
they were closing down shop.
The update on the analysis of samples from the National Children’s Study was a confusing one.
The lab had completed the analyses of the bloodspots, the cord blood and the maternal blood,
but there were some unexpected results, especially for the bloodspots, and a concern that
there had been mercury contamination that had occurred in the collection or processing. As
MN Biomonitoring was not sure what to make of that, the NCS was providing assistance with
40
investigating whether contamination could have occurred along the way. They had already
provided the mothers’ demographic and survey data, and Jessica stated that they were
continuing their analysis.
Lastly, Jessica reminded the panel of the current MN FEET project, and how that project would
fit into these studies. With input from the panel and the smaller mercury studies, the
biomonitoring team had designed MN FEET to help answer some of the same questions. MN
FEET’s ancillary study would also add a greater sample size (roughly 300) for bloodspot and
cord blood comparison, which was a validation that had been requested nationwide. Since the
choice had been made to collect bloodspot samples from the subset of women with higher cord
blood levels, we would have a high number of detectable bloodspot samples that would enable
this validation to be done. MN FEET’s large sample size (600 cord blood, urine and 300
bloodspots) would also lead to the ability to characterize newborn exposures and the sources
of the exposure. Even though it currently had a metro area focus of certain communities, she
hoped to expand to non-metro parts of the state in the future. The team had designed the
study to look at disparities in exposure in four groups, Hmong, Latina, Somali and White
women.
Discussion
The following question was asked of the panel: Does the panel have any recommendations for
additional analyses, data interpretation, or next steps? Greg Pratt inquired about the
systematic difference between the bloodspot and blanks, with the mercury in the bloodspot
sometimes being lower than the mercury found in the bloodspot blanks. What was the
magnitude of the mercury levels in the blanks or the difference between the levels? Jessica
responded that in the TIDES Study, they had not seen any mercury levels in the blanks. Betsy
Edhlund added that most of the time, the blanks had lower levels of mercury than the
bloodspots and that the Riverside study had levels of 0.7-1.4 µg/L mercury in blanks.
Bruce Alexander asked about the adjustments done for the blanks in the sample, how did
Jessica design this? Jessica responded that Utah had done this already and that had been the
method she was using. Utah had mercury in their blank samples as well, and they had been
subtracting them, as there could have been contamination.
Bruce also asked for clarification from Jessica as to whom she was working with to recruit
participants for the MN FEET study, as that had been a challenge for the RBS project. Jessica
described the design of MN FEET as going through prenatal clinics, working with the research
arms of Health Partners and West Side Community Health Services, with five or six prenatal
clinics, and they would actually be pulling from their patient lists and doing a combination of a
phone call and a letter. To optimize recruitment, we were pairing with a community outreach
effort and some provider education, so it seemed the most powerful way for a woman to hear
about this would be through their provider. This was a change from our original plan of
recruiting directly in the clinics, but that had to be changed due to limited appointment time at
the clinic. So now it would be a random sample of patients being sent a letter and then having a
follow-up phone call.
Bruce asked whom the community outreach was being done through. Jessica responded that
the team had worked with Hmong, Latina, Somali and White women, and the groups we had
41
been talking with were St. Paul Ramsey Public Health, due to their great community outreach to
those groups. They had offered to connect us with radio shows and other groups. We had also
worked with the research arm of West Side Community Health Services, which was a group of
community-based researchers from Somalia, Latina and Hmong communities (SoLaHmo). Pat
McGovern added that it sounded as though the team had done a nice job of getting the right
groups to the table.
Biomonitoring Updates
Carin Huset described the updates to the PFC laboratory method. Since PFNA was a new
analyte (it was not included in previous studies), the previous method required revalidation.
During this revalidation, she encountered problems with reproducibility and chose to make
changes to the method to make it more reproducible, more robust, and higher throughput.
The new method (described on page 11 of the February 10, 2015 advisory panel background
material book) increases the number of unknown samples that could be analyzed from about
30 to about 60 at a time. Some of the changes she described: previously they had stable isotope
labeled internal standards for six of the seven analytes, now they have internal standards for all
analytes; they decreased the volume of sample used for the analysis.
The new method of sample preparation involves protein precipitation to remove proteins that
were causing problems with reproducibility and robustness. When the proteins are
precipitated, the sample is a lot simpler to work with. The validation includes pooled quality
controls that were measured repeatedly during validation and then pairs (of low and high
standards) that are analyzed with each batch of unknowns; this is what the CDC does for their
biomonitoring standards and what MDH has done for their CDC- Laboratory Response Network
Program.
Another change from 2010 is increased options for proficiency testing. There are more analytes
available for external proficiency testing; now there are five analytes instead of just two, which
gets beyond one of our previous limitations. Carin explained that proficiency testing is when an
external company would develop samples, send them to MDH (and other labs) and MDH (and
the other labs) send back results for the analysis. The external company analyzes the results of
all the participating labs and provides a report back to the participating labs on their
performance.
Carin also described the lengthy process of testing the methods for PFCs due to their presence
in many consumer products and lab products. We had to make sure there were low levels in
background before we could begin. We also had to be sure the results determined by the new
method were comparable to results from previous studies (which were originally analyzed with
the old method). They pulled samples from the 2008 study; they were comparable within+-20%
(the number used to assess duplicates).
Carin reported on the current status of 490 samples completed for the GLRI study, and 150
samples analyzed for the PFC3 project.
42
Geary Olsen asked Carin to go back to the PFOA graph and asked if there were more samples
than was shown. Carin responded that they only had been given three samples three times a
year. Gary wondered if Carin knew why they did not have samples that had PFOA at levels more
similar to what is observed in the general population, because the numbers shown were quite
high and it made it too easy to reach the middle. Carin was not sure if they were historical or
spiked samples, but she agreed that the levels were high relative to what is observed, and many
of the samples needed to be diluted by MDH in order to report the values.
Geary Olsen asked whether PFBA was one of the analytes looked at, but not shown in the
external QC. Carin explained that they were in the internal, but not external QC. She added that
as far as she could tell, MDH’s Public Health Laboratory was one of the few labs looking for
PFBA at this time, and that she had been asked to describe the process to other labs in the
country.
Hearing no other questions for Carin, Jean announced that during the refreshment break, she
would give an update on progress with our work plan for Sustaining Minnesota Biomonitoring
by playing a portion of Jessica Nelson’s interview with Commissioner Ehlinger about MN
Biomonitoring on A Public Health Journal.
East Metro Community Cancer Report
Kenneth Adams, Minnesota Cancer Surveillance System, presented a data update to the 2007
East Metro Community Cancer Report. Full background materials can be found on pages 17-38
of the February 10, 2015 Advisory Panel Meeting Book. Kenneth referred the panel to the
background and methods on pages 23-26; results in text form on pages 26-29, and specifically,
the county level results on pages 30-33, upon which his presentation was focused.
Kenneth gave a tutorial on understanding cancer statistics, including case counts, rates, and
age-adjusted rates. The number of individuals with cancer in Minnesota is increasing each year,
which is consistent with many peoples’ perception from their daily lives. One reason cancer is
becoming more common is that the state’s population is increasing, especially in the suburbs.
For example, the combined population of Washington and Dakota Counties grew from 390,000
to 650,000, or 68%, between 1988 and 2012. Epidemiologists often express cancer occurrence
in terms of crude rates as a way to account for changes in population size over time, and
differences in population size between places. A second reason more cancers are occurring is
that the population is aging, and cancer was often a disease of aging. In 1990, a typical adult in
Washington or Dakota Counties was 30-35 years old, but in 2010, the typical age has increased
to 50-55 years. Epidemiologists age-adjust rates to take account of both the increase in
population and the aging of the population; this allows rates to be compared over calendar
time. In contrast with cancer counts in Washington and Dakota Counties, age-adjusted cancer
rates have been fairly steady over the past 25 years (based on data through 2012).
Kenneth then presented results from new analyses that MCSS has prepared. These new
analyses will be compiled into a new report, which will update the 2007 MCSS Report: Cancer
Incidence in Dakota and Washington Counties. The new data update replicates the key results
of the 2007 report, and adds new results based on data collected by the cancer registry up
43
through 2012. The data update is based on indirect standardization, which is the same standard
epidemiologic methodology used in the 2007 report. Kenneth reviewed the statistics and
results presented in the report.
The methods used were indirect age standardization for the observed population, county or
community. This was the aggregated or the observed number of cases overall, in an area, in our
registry for this time period. The reference population or comparison population, for county
level analyses, was the State of Minnesota. A key statistic was the number of “expected” cases,
which was the number of cases that would be expected if the observed population had the
cancer experience of the reference population. This statistic was compared with the observed
cases, the number of cases that actually had occurred in the county or community.
Kenneth then reviewed the tables starting on page 30 of the background materials book,
clarifying the following definitions for understanding the report update results:
Cases Observed was the MCSS registry count of newly diagnosed cancers among Minnesota
residents. Cases Expected was a modeled estimate of the number of cases that would occur if
the observed population (the county or the community) had the same cancer rates as the
reference population (e.g., the State of Minnesota). The Observed-to-Expected Difference
estimated the potential public health impact, and was similar to risk difference; the Observedto-Expected Ratio corresponded to relative risk; the 95% Confidence Interval was the range of
plausible estimates for the observed-to-expected ratio.
The county-level analysis included over 200 separate statistical results. The following steps had
been taken to identify or characterize unusually high occurrence of cancer: the initial step was
to identify those results that were statistically significant. Further steps were to consider or
evaluate the consistency over calendar time and between males and females; the magnitude of
estimates, as differences and as ratios; the confidence interval width, the potential variability of
the estimates.
The results represented or characterized the overall cancer experience of the observed
population over the calendar interval evaluated. The results represented the combined effect
of all known and unknown environmental, genetic and biological factors influencing cancer risk,
including chance and random variability. They were not specific to any putative environmental
exposure. Kenneth felt the results answered the question of whether the cancer experience of
the community was unusual in the every-day meaning of the term “unusual”. They also
provided information and could educate on the nature of cancer—that it was very common and
the occurrence was highly variable over time and place.
Kenneth pointed out that most results are not consistent across calendar periods. An exception
was Washington County males, where the number of observed lung cancer cases was less than
expected in both calendar periods, and Dakota County females, where the number of observed
breast cancer cases was greater than expected in both calendar periods. He also noted that
among Minnesota’s 87 counties, the age-adjusted breast cancer rates ranged from 63 to 161
cases per 100,000 annually, from 2003 to 2012.The numbers of new cases were higher than
expected for some cancer types and lower than expected for others. Most did not differ. He
explained that this was not surprising; cancer rates were known to vary considerably over time
44
and place. Few of the differences were consistent over calendar time or between males and
females, and in most, but not all analyses the magnitude of difference was not large. Overall,
these results did not suggest that occurrence of cancer in Washington or Dakota Counties was
unusually high. A notable exception was breast cancer. More newly diagnosed breast cancer
cases were observed in Washington and Dakota Counties than expected. Among 87 counties in
Minnesota, Washington County had the 8th highest rate of female breast cancer and Dakota
County had the 15th highest rate, over the time period 2003 to 2012.
Discussion
Greg Pratt commented that cancer rates vary in time and space and he asked Kenneth how he
had calculated the expected rates. Kenneth responded by reviewing the indirect
standardization methodology used in the analysis.
Fred Anderson asked whether the report was online, and what the next steps were for
publishing. Kenneth replied that he, Alan Bender and Jean Johnson would discuss it. Jean noted
that she would like it to be ready when the PFC biomonitoring results were distributed in the
spring, so that participants could have their questions answered.
Pat McGovern wondered about the biological plausibility that PFCs cause cancer. Bruce
Alexander responded that bladder, pancreas and liver cancers have come up in small studies,
and kidney cancer has not really borne out. He added that in the Cottage Grove plant of
workers with high exposure, there were no obvious cancer results.
Jill Heins-Nesvold asked how the 8 zip codes had been selected in Washington County. Jean
answered that they had been based on the PFC plume, it included every zip code in the plume.
Jill added that one additional explanation for the breast cancer result could be that Washington
County was the most affluent county in Minnesota. She wondered whether we could look at
the percentage of females who had gotten mammograms to see if there was an interesting
correlation.
Pat McGovern noted that the community might want more information on known factors
associated with breast cancer. The residents may have questions they would want to ask their
providers, so the information would need to be put into context. Lisa Yost agreed and
wondered whether there would be a follow up or whether we would look further into other risk
factors? She questioned whether there was the ability to drill down in the data? Kenneth
Adams answered that they could not drill down or explain away; breast cancer was somewhat
modifiable. People have been educated to take steps to minimize the cancer risk, such as eating
more healthy foods and getting a mammogram. Fred Anderson wondered if we could somehow
adjust or account for the confounder of access to care. Margee Brown, Minnesota Cancer
Surveillance System, responded that delaying childbearing also would need to be considered.
Jean Johnson asked the panel for the best way to present this information to the public, the key
messages.
Geary Olsen noted that a study had been done for the plume in Washington County, for issues
revolving around eight communities. 3M and the University of Minnesota published
occupational cohort, cancer incidence and mortality data for the plant and they found no
association with breast cancer in 800 women. The C8 Science Panel out of Washington found
45
that breast cancer was not an issue. IARC found possible for testicular and kidney cancers.
Geary recommended that we needed to understand the Dakota and Washington County
mortality rates before we released the information to the public. Bruce Alexander commented
that if you were to go to the community to say that there was higher breast cancer in Oakdale;
they would think it was because of the PFCs, so we would have to put it in context of
information on PFCs and breast cancer.
Jean Johnson asked Fred Anderson if the county had information on risk factor data, such as
delayed childbearing or access to care? Fred thought that the planners had some information.
He added that there were also other health cluster concerns in Washington County, such as
brain cancer in Stillwater; various cancers in a mining area; if there might be a contextual
component, then we would need to be ready to discuss future concerns.
Pat McGovern asked whether Jean wanted any input from the panel on the message as it was
being written. Bruce suggested we talk to the breast cancer prevention program at MDH.
Prostate cancer might be another one. He added that Kenneth’s initial graphs showing
population over time were important.
Lisa Yost wondered if we were attempting to link with exposure from biomonitoring? Jean
Johnson answered that were not. Geary Olsen added that C8 study looked at multiple
exposures of PFCs in a fluoropolymer plant, included a known kidney cancer toxicant (PTFE) and
found no excess kidney cancer in 3M plant.
Greg Pratt referred to the graph with the counties--those with the highest breast cancer rates—
did we know anything about those counties? Kenneth Adams answered that they are rural with
smaller populations.
Jill Heins-Nesvold said that the population doubling had been a good point. The population in
and out was changing and the population brought in and took out cancers with them. Kenneth
Adams added that cancer followed where the person moved. Jean Johnson said that in the
past, in St. Louis Park, MDH looked at whether the Jewish population could have had an
influence on the breast cancer rates, due to known higher rates in Jewish women. She also
mentioned looking at the smoking prevalence to put the lung cancer findings into context.
Fred Anderson commented that this was really helpful from a Washington County perspective.
Many departments get questions when reports come out, so programs will need some talking
points. Kenneth offered, if anyone requested, to provide the zip code results.
Tracking Updates
Matthew Montesano gave a portal demonstration of the new Interactive Asthma
Hospitalization map, part of a new interactive portal system. Matthew explained that it was
mobile compatible, which affected about 20% of the traffic to the portal. He described the new
system as integrated with an intuitive interface, there was county and zip code data on the
same page, and that it was more efficient to build, maintain and improve. It was a simplified
design, using the principle that if you had to explain something, it was not very good. Other
46
updates could be found on pages 43-47 of the February 10, 2015 Advisory Panel background
book.
Geary Olsen asked if there had been an example of data on the portal leading to a request of
the Minnesota Department of Health to do a cluster investigation, with MDH subsequently
agreeing to do so? Matthew added that he did not know of anything directly coming from
someone searching and then contacting the department. It does happen when the media
highlights an issue.
Greg Pratt asked if Matthew knew a breakdown of how people were getting to the portal. Lisa
Yost wondered how easy it was for people to have direct contact with the program about
something they saw on the portal. Matthew replied that they received about one to two emails
per month. Bruce Alexander suggested that the asthma hospitalization rates by zip code would
benefit from adding the time period to the legend. Dan Tranter asked how many unique visitors
the portal had per year. Mathew responded that unique visits were not tracked anymore;
google changed their analytics to sessions, so 3000 sessions per month was about 2000 users
per month. Jean Johnson added that they now had a new audience of portal users; 2/3 of the
portal users were academics due to recent outreach.
East Metro PFC3 Biomonitoring Project Update
Christina Rosebush presented an update on the progress of the PFC3 Biomonitoring project.
Background materials were found on pages 47-51 of the February 10, 2015 Advisory Panel
meeting book.
She reviewed the key questions the study had been designed to ask:
•
•
•
Had PFC levels continued to decline in long-term East Metro residents?
In new Oakdale residents, were PFC levels comparable to U.S. general population?
Was there an association between length of residence in Oakdale since October 2006
and PFC levels?
Christina Rosebush gave an update on recruitment, which began in February of 2014 and was
completed in January of 2015. There were three study groups recruited, the Original Cohort
group, New Residents and New Renters.
There were recruitment delays, Christina explained, due to the two-step process for recruiting
the New Residents. First, a household survey was sent to identify eligible individuals, then
participants were randomly selected and invited to participate. Additionally, a New Renters
sample was added in August, first identifying the sampling frame through Washington County
Housing and Redevelopment Authority, then repeating the two-step process of a household
survey and individual recruitment.
She continued that the participation from the New Resident groups (49%) was, as expected,
lower than the response from the Original Cohort (89%). They had not participated previously
and many were unaware of PFC history in the East Metro.
47
Age was significantly different among the 3 groups: Original Cohort, New Residents, and New
Renters. Within the New Residents group, Renters were slightly older–one of the HRA
properties was a Senior Living facility. Length of residence was significantly longer for The
Original Cohort, which was expected because eligibility for these groups was based on
residence (<1/2005 for Original Cohort).
Christina noted that race/ethnicity was significantly different among the 3 groups. Looking only
at New Residents and New Renters, there was still a significant difference in Race/ethnicity.
Most of the Non-white New Renters were Black/African-American or African. Among all New
Residents, the other Non-White groups were White/Hispanic and Asian.
Christina continued that Income was significantly different among the three groups. It was also
significantly different when looking just at New Residents and Renters. HRA records were used
to identify Renters, and there were income requirements for renting through the HRA. Income
requirements varied by HRA property.
The analysis plan for the Original Cohort, Christina explained, would be to compare their results
to the NHANES 2011-12 data and compare the percent change for PFOS, PFDA and PFHxS since
the 2008 and 2010 Minnesota Department of Health PFC projects. For New Residents, she
would compare the levels to NHANES data from 2011-2012 and also analyze the association
between length of residence since the water treatment intervention and their PFC levels.
Regarding half-lives of these chemicals, Christina mentioned plans to do an elimination rate
calculation to check that it was consistent with published ½ lives. It would not be a true half-life
because we did not have control over all sources of exposure. At some of the low PFC levels
seen even in the Original Cohort, other sources of exposure might be significant (unlike
occupational studies where PFC levels were so high that other sources were not as significant).
Christina noted some possibilities for the new renter analysis. Since the new renters were
limited by a small sample size of 19, she proposed to test PFC levels in New Residents versus
New Renters. If there were no differences found, they could be grouped together in analysis.
Alternatively, all the results could be presented separately due to significant demographic
differences. Another possibility would be to take the six Renters identified through the Oakdale
water billing records (now part of the New Residents) and add them to the New Renters group
for sensitivity analysis.
Christina posed the following questions for the panel:
•
Given that we only had a small number of Renters, should New Renters be grouped with
New Residents for analysis?
• Did the Panel have other comments/recommendations for the analysis plan?
Discussion
Bruce Alexander asked what the real question was, Renters or socioeconomic status? Christina
explained that the Original Cohort had excluded Renters, so that was why they had been added.
Geary Olsen commented that MN Biomonitoring could not analyze Renters; there were only 4
males in the group and they could become identifiable. He was not sure how to get around it,
other than to not group them by gender and collapse them into a non-gender group. He added
48
that 19 people was a small sample size regardless. Bruce Alexander added that you could argue
that because we have included Renters, representative of a New Resident group, that could
lead to an argument to look at them combined. You could look at it by income and race in one
group. Geary Olsen commented that New Residents and New Renters all were on municipal
water in Oakdale, so the only difference was demographics. The one exposure in common was
Oakdale water, so collapsing them together made sense. Bruce wondered if Christina had
specific information on non-respondents, to which Christina replied that she did not. Greg Pratt
proposed looking at the significant differences and lumping them together if there were none.
The question he would ask was what if there were differences in two groups for some PFCs. Do
the power calculation to see what difference could be detected with 19, if that was enough.
Christina reviewed the power calculation; GM differences between 1-2 ug/mL could be
detected, but the question of what difference was clinically significant remains. What would we
do if we saw significant differences? Was there an argument to keep the groups separate? Lisa
Yost asked what the main question was--that now that everyone was on the same water
supply, was it the difference between Renters and homeowners-or compared to NHANES. She
suggested combining the groups and looking at PFC differences by income.
Jeanne Ayers discussed the fact the original study had not brought in the racial inequity piece –
the Renter population was added as a strategy to get to a more diverse sample. The
income/race analysis was an important one—it was not about renting or not, but to get at the
socioeconomic factors. For Advancing Health Equity—if everything was done based on home
ownership, people are left out. Only 20-25% of people in minority groups are homeowners.
Christina clarified that this was not a sample of all renters in Oakdale; it was just renters from a
few properties. She added that it was more difficult to reach out to renters. Jeanne Ayers
replied that it was not as pure a design, not perfect, but its intentionality was improved.
Lisa Yost wondered whether participants were compensated, to which Christina responded that
they received a $25 gift card. Bruce Alexander asked what the comparison would be in the
results that would be returned to participants, to which Christina answered NHANES.
Geary Olsen asked what percent of the NHANES sample could have been from Washington
County. Christina said we do have estimates regarding the inclusion of Washington County
residents in NHANES 2011-12 that she could compile for the next panel meeting.
Jean Johnson asked, based on the new information the panel had just seen, whether the team
had answered the rental/race disparity question. Jeanne Ayers referred to the first question the
panel had been asked; (Given that we only had a small number of Renters, should New Renters be
grouped with New Residents for analysis?) and believed we should combine the group for analysis
and look at demographic differences rather than use a separate “renter” group.
Melanie Ferris questioned how much work was done within MDH to communicate consistently
in getting results to a larger community. She wondered if MDH could coordinate information to
the same common residents, packaging the communications that come from multiple sections
of the Minnesota Department of Health. She was concerned about the broader questions the
community residents might have beyond PFCs. Jim Kelly said the Minnesota Department of
Health already worked together to have consistent information on PFC technical issues, but
49
more work could be done on the content for breast cancer, or they could consider other
programs. Jean Johnson added that MDH works with MPCA on air pollution and health.
Ongoing PFC Study in the East Metro
Jean Johnson asked the panel whether they would recommend any additional PFC
biomonitoring in the East Metro area in the future, given that results of PFC3 were not yet
available. Hearing no recommendation, she then asked if they would recommend PFC
biomonitoring in another community. There was no recommendation.
Future Meeting Topics
Pat McGovern invited the panel members to recommend topics for exploration and discussion
at future meetings. Jill Heins Nesvold mentioned ambient air quality study with cancer and also
other endpoints as cancer would be only one result. Jean Johnson stated that the report on
respiratory/heart disease and air in communities should be ready by June.
Pat McGovern added that she was interested in the focus of health equity and how it might
apply to this group, other than the example we discussed at today’s meeting. An example
would be the air pollution exposure and health outcomes, in terms of health disparities. Greg
Pratt noted that he was interested in this topic as well. He added that he has a paper on
inequities in risk from air pollution exposure, looking at modeled risks for The International
Society of Exposure Science and would be willing to talk to the group about his work. Jean
Johnson referenced the work of Julian Marshall, an interesting study being done at the
University of Minnesota. Jeanne Ayers added that Greg Pratt’s work on disparity in air pollution
exposure or exposure work in general could be used to raise questions or could point the group
to generate areas that advance health equity and need additional study.
Geary Olsen said pharmacokinetics was the intermediary that was being forgotten if you only
study biomonitoring and health effects. Regardless of the compound being studied, the
underlying pharmacokinetics of anything being measured is very important. He continued that
in order to understand health effects, we need to go through pharmacokinetics first to see the
underlying reasons why we might see associations between biomonitoring and health. He
added that he had some presenters in mind.
Bruce Alexander would be interested in reviewing the story of the MDH AND University of
Minnesota NE Minneapolis community vermiculite study. Greg Pratt added the University of
Minnesota NE Minnesota mining study.
Pat McGovern asked whether biomonitoring could compliment a Deanna Scher study of well
water testing in Dakota County, maybe looking at other metals. She thought there might be
implications for biomonitoring to complement that work. She also mentioned that Jill Prescott
was interested in manganese as well. Jim Kelly responded that he has had discussions with
Dakota County in order to understand what was in ground water and what messaging residents
get regarding water testing that could help influence them to get testing or take action when
50
they received results. He offered to come back to the group when he was further along. He
added that 20% of the Minnesota population was on private well water that was not regulated.
Jeanne Ayers noted that the Minnesota Department of Agriculture had private well sampling
projects and was interested in nitrates. Cathy Villas-Horns replied that tens of thousands of
wells were being sampled and that she would contact some potential speakers within MDA that
could attend a future meeting to speak on this project. Jeanne commented that there was a
political divide on agricultural use of nitrates and health risk and that better scientific
information was needed.
Jean Johnson added that Minneapolis was monitoring Polycyclic aromatic hydrocarbons (PAHs).
Greg Pratt thought that the sampling would be completed in June; then there would be a lab
analysis, so he could share a video about the efforts.
New Business
There was no new business.
Audience Questions
There were no audience questions.
Adjournment
The meeting was adjourned at 4:00 pm. The next Advisory Panel meeting will be held on June 9,
2015 from 1:00–4:00 p.m. at the American Lung Association in Minnesota.
51
Environmental Health Tracking and Biomonitoring
Advisory Panel Roster
As of March 1, 2015
Bruce Alexander, PhD
School of Public Health
University of Minnesota
Environmental Health Sciences Division
MMC 807 Mayo
420 Delaware Street SE
Minneapolis, Minnesota 55455
612-625-7934
[email protected]
At-large representative
Fred Anderson, MPH
Washington County
Dept. of Public Health & Environment
14949 62nd St N
Stillwater MN 55082
651-430-6655
[email protected]
At-large representative
Alan Bender, DVM, PhD
Minnesota Department of Health
Health Promotion & Chronic Disease
Division
85 East 7th Place
PO Box 64882
Saint Paul, MN 55164-0882
651-201-5882
[email protected]
MDH appointee
Melanie Ferris
Wilder Foundation
451 Lexington Parkway N
St. Paul, MN 55104
651-280-2660
[email protected]
Nongovernmental organization
representative
Thomas Hawkinson, MS, CIH, CSP
Toro Company
8111 Lyndale Avenue S
Bloomington, MN 55420
[email protected]
952-887-8080
Statewide business org representative
Jill Heins Nesvold, MS
American Lung Association of
Minnesota
490 Concordia Avenue
St. Paul, Minnesota 55103
651-223-9578
[email protected]
Nongovernmental organization
representative
Pat McGovern, PhD, MPH
School of Public Health
University of Minnesota
Environmental Health Sciences Division
MMC Mayo 807
420 Delaware St SE
Minneapolis MN 55455
612-625-7429
[email protected]
University of Minnesota representative
Geary Olsen, DVM, PhD
3M Medical Department
Corporate Occupational Medicine
MS 220-6W-08
St. Paul, Minnesota 55144-1000
651-737-8569
[email protected]
Statewide business organization
representative
52
Steven Pedersen, MPH
8403 Mississippi Boulevard NW
Coon Rapids, MN 55433
612-850-1058
[email protected]
Minnesota Senate appointee
Gregory Pratt, PhD
Minnesota Pollution Control Agency
Environmental Analysis & Outcomes
Division
520 Lafayette Road
St. Paul, MN 55155-4194
651-757-2655
[email protected]
MPCA appointee
Andrea Todd-Harlin, MS
Medical Research Advisors
1491 McCarthy Road
Eagan, MN 55121
651-341-3444
[email protected]
MN House of Representatives appointee
Cathy Villas-Horns, MS, PG
Minnesota Dept. of Agriculture
Pesticide & Fertilizer Management
Division
625 Robert Street North
St. Paul, Minnesota 55155-2538
651-201-6697
[email protected]
MDA appointee
Lisa Yost, MPH, DABT
ENVIRON International Corporation
333 West Wacker Drive, Suite 2700
Chicago, IL 60606
Local office
479 Iglehart
St. Paul, MN 55103
Phone: 651-225-1592
Cell: 651-470-9284
[email protected]
At-large representative
53
Biographical sketches of advisory panel members
Bruce H. Alexander is a Professor in the Division of Environmental Health Sciences at the
University of Minnesota’s School of Public Health. Dr. Alexander is an environmental and
occupational epidemiologist with expertise in cancer, reproductive health, respiratory disease,
injury, exposure assessment, and use of biological markers in public health applications.
Fred Anderson is an epidemiologist at the Washington County Department of Public Health and
Environment and has over 30 years of public health experience. He holds a Master’s of Public
Health (MPH) in environmental and infectious disease epidemiology from the University of
Minnesota and is a registered environmental health specialist. For over 20 years, he has led
county-wide disease surveillance and intervention programs, including numerous
multidisciplinary epidemiologic investigations.
Alan Bender is the Section Chief of Chronic Disease and Environmental Epidemiology at the
Minnesota Department of Health. He holds a Doctor of Veterinary Medicine degree from the
University of Minnesota and a PhD in Epidemiology from Ohio State University. His work has
focused on developing statewide surveillance systems, including cancer and occupational
health, and exploring the links between occupational and environmental exposures and chronic
disease and mortality.
Melanie Ferris, MPH, is a Research Scientist at Wilder Research, a nonprofit research
organization based in St. Paul, Minnesota. She conducts a variety of program evaluation and
applied research projects focused primarily in the areas of public health and mental health. She
has worked on a number of recent projects that focus on identifying disparities across
populations and using existing data sources to develop meaningful indicators of health and
wellness. Examples of these projects include a study of health inequities in the Twin Cities
region related to income, race, and place, development of a dashboard of mental health and
wellness indicators for youth living in Hennepin County, and work on local community health
needs assessments. She has a Master’s of Public Health degree in Community Health Education
from the University of Minnesota’s School of Public Health.
Tom Hawkinson is the Corporate Environmental, Health, and Safety Manager for the Toro
Company in Bloomington, MN. He completed his MS in Public Health at the University of
Minnesota, with a specialization in industrial hygiene. He is certified in the comprehensive
practice of industrial hygiene and a certified safety professional. He has worked in EHS
management at a number of Twin Cities based companies, conducting industrial hygiene
investigations of workplace contaminants and done environmental investigations of subsurface
contamination both in the United States and Europe. He has taught statistics and mathematics
at both graduate and undergraduate levels as an adjunct, and is on the faculty at the Midwest
Center for Occupational Health and Safety A NIOSH-Sponsored Education and Research Center
School of Public Health, University of Minnesota.
Jill Heins Nesvold serves as the Director of the Respiratory Health Division for the American
Lung Association in Iowa, Minnesota, North Dakota, and South Dakota. Her responsibilities
include program oversight and evaluation related to asthma, chronic obstructive lung disease
(COPD), lung cancer, and influenza. Jill holds a Master’s Degree in Health Management and a
short-course Master’s Degree in Business Administration. Jill has published extensively in a
variety of public health areas.
Pat McGovern is a Professor in the Division of Environmental Health Sciences at the University
of Minnesota’s School of Public Health. Dr. McGovern is a health services researcher and nurse
with expertise in environmental and occupational health policy and health outcomes research.
She served as the Principal Investigator for the National Children’s Study (NCS) Center serving
Ramsey County, one of 105 study locations nationwide. The NCS was the largest, long-term
study of children’s health and development in the US and the assessment of environmental
exposures will include data collection from surveys, biological specimens and environmental
samples.
Geary Olsen is a corporate scientist in the Medical Department of the 3M Company. He
obtained a Doctor of Veterinary Medicine (DVM) degree from the University of Illinois and a
Master of Public Health (MPH) in veterinary public health and PhD in epidemiology from the
University of Minnesota. For 27 years he has been engaged in a variety of occupational and
environmental epidemiology research studies while employed at Dow Chemical and, since
1995, at 3M. His primary research activities at 3M have involved the epidemiology,
biomonitoring (occupational and general population), and pharmacokinetics of
perfluorochemicals.
Steven Pedersen is a retired Environment, Health, and Safety (EHS) scientist who worked for
BAE Systems in Fridley, MN. He completed his Master’s in Public Health at the University of
Minnesota, with a specialization in environmental health. He has thirty-five years’ experience
working on EHS issues; focusing on environmental compliance and the development and
implementation of a management system compliant with the requirements of the international
standards. He has worked in EHS project management at a number of aerospace companies in
Minnesota, Washington, and California. He worked on environmental legislative and regulatory
issues and is an expert on the requirements of the Toxic Substances Control Act as it affects
article-manufacturing companies. He was the project manager implementing an enterprisewide Occupational Safety, Health, and Environment (OSHENs) illness & injury data-management
system. Recently he was a Governor-appointed member, representing the business community,
of the State's Clean Water Council.
Gregory Pratt is a research scientist at the Minnesota Pollution Control Agency. He holds a
Ph.D. in Plant Physiology from the University of Minnesota, where he worked on the effects of
air pollution on vegetation. Since 1984 he has worked for the MPCA on a wide variety of issues
including acid deposition, stratospheric ozone depletion, climate change, atmospheric fate and
dispersion of air pollution, monitoring and occurrence of air pollution, statewide modeling of
air pollution risks, and personal exposure to air pollution. He is presently cooperating with the
Minnesota Department of Health on a research project on the Development of Environmental
Health Outcome Indicators: Air Quality Improvements and Community Health Impacts.
55
Andrea Todd-Harlin is an epidemiologist with 15 years’ experience in both the public and
private sectors. She holds a Master’s of Science in Environmental Epidemiology & Policy from
the London School of Hygiene and Tropical Medicine and a Bachelor’s of Science in Health &
Wellness from the University of Minnesota. Andrea began her career at the Minnesota
Department of Health in the Chronic Disease and Environmental Epidemiology section, where
she worked on grants researching serious traumatic work-related injury and childhood asthma.
She then moved into applied practice, serving as the Director of Research and Education at the
private medical practice Sports and Orthopaedic Specialists. Andrea has also served as adjunct
faculty at St. Catherine University and Argosy University teaching microbiology, biostatistics and
epidemiology and risk management. She currently operates her own medical research
consulting firm, Medical Research Advisors.
Cathy Villas Horns is the Hydrologist Supervisor of the Incident Response Unit (IRU) within the
Pesticide and Fertilizer Management Unit of the Minnesota Department of Agriculture. Cathy
holds a Master of Science in Geology from the University of Delaware and a Bachelor of Science
in Geology from Carleton College and is a licensed Professional Geologist in MN. The IRU
oversees or conducts the investigation and cleanup of point source releases of agricultural
chemicals (fertilizers and pesticides including herbicides, insecticides, fungicides, etc. as well as
wood treatment chemicals) through several different programs. Cathy has worked on complex
sites with Minnesota Department of Health and MPCA staff, and continues to work with
interagency committees on contaminant issues. She previously worked as a senior
hydrogeologist within the IRU, and as a hydrogeologist at the Minnesota Pollution Control
Agency and an environmental consulting firm.
Lisa Yost is a Principal Consultant at ENVIRON, an international consulting firm. She is in
their Health Sciences Group, and is based in Saint Paul, Minnesota. Ms. Yost completed her
training at the University of Michigan’s School of Public Health and is a board-certified
toxicologist with expertise in evaluating human health risks associated with substances in soil,
water, and the food chain. She has conducted or supervised risk assessments under CERCLA,
RCRA, or state-led regulatory contexts involving a wide range of chemicals and exposure
situations. Her areas of specialization include exposure and risk assessment, risk
communication, and the toxicology of such chemicals as PCDDs and PCDFs, PCBs,
pentachlorophenol (PCP), trichloroethylene (TCE), mercury, and arsenic. Ms. Yost is a
recognized expert in risk assessment and has collaborated in original research on exposure
issues, including background dietary intake of inorganic arsenic. She is currently assisting in a
number of projects including a complex multi-pathway risk assessment for PDDD/Fs that will
integrate extensive biomonitoring data collected by the University of Michigan. Ms. Yost is also
an Adjunct Instructor at the University of Minnesota’s School of Public Health.
56
Staff Biosketches
Kenneth F Adams, PhD, is an epidemiologist with the Minnesota Cancer Surveillance System
(MCSS), Minnesota’s central cancer registry. His day-to-day work includes estimation of cancer
rates, performance of record linkages between MCSS and other data, responding to citizen
cancer concerns, and data collection for a screening colonoscopy research study. He was
formerly a postdoctoral fellow in the US National Cancer Institute Division of Cancer
Epidemiology and Genetics, and a research investigator at HealthPartners Institute. He received
a PhD in epidemiology from the University of Washington in 2003.
Wendy Brunner, PhD, serves as surveillance epidemiologist for the MDH Asthma Program since
2002, and joined Minnesota’s Environmental Public Health Tracking and Biomonitoring Program
(MN Tracking) program on a part-time basis in fall 2009. Previously, she worked on occupation-al respiratory disease studies for MDH. She has a Master’s Degree in Science and Technology
Studies from Rensselaer Polytechnic Institute and a Master’s Degree in Environmental and
Occupational Health from the University of Minnesota. She received her doctorate in the
Division of Epidemiology and Community Health at the University of Minnesota.
Betsy Edhlund, PhD, is a research scientist in the Environmental Section of the Public Health
Laboratory at the Minnesota Department of Health. She works in the metals laboratory
developing methods and analyzing samples for both biomonitoring programs and emergency
response. Betsy received her PhD in chemistry from the University of Minnesota where her
research focused on the photochemistry of natural waters.
Carin Huset, PhD, has been a research scientist in the Environmental Laboratory section of the
MDH Public Health Laboratory since 2007. Carin received her PhD in Chemistry from Oregon
State University in 2006 where she studied the fate and transport of perfluorochemicals in
aqueous waste systems. In the MDH PHL, Carin provides and coordinates laboratory expertise
and information to program partners within MDH and other government entities where studies
require measuring biomonitoring specimens or environmental contaminants of emerging
concern. In conjunction with these studies, Carin provides biomonitoring and environmental
analytical method development in support of multiple analyses.
Jean Johnson, PhD, MS, is Program Director/Principal Investigator for MN Tracking. Dr. Johnson
received her Ph.D. and M.S. degrees from the University of Minnesota, School of Public Health
in Environmental Health and has 25 years of experience working with the State of Minnesota in
the environmental health field. As an environmental epidemiologist at MDH, her work has
focused on special investigations of population exposure and health, including studies of
chronic diseases related to air pollution and asbestos exposure, and exposure to drinking water
contaminants. She is currently an adjunct faculty member at the University of Minnesota’s
School of Public Health.
Tess Konen, MPH, graduated from the University of Michigan’s School of Public Health with a
Master’s Degree in Occupational Environmental Epidemiology. She completed her thesis on the
effects of heat on hospitalizations in Michigan. She worked with MN Tracking for 2 years as a
CSTE Epidemiology Fellow where she was project coordinator for a follow-up study of the
57
Northeast Minneapolis Community Vermiculite Investigation cohort. She currently is an
epidemiologist working on birth defects, pesticides, and climate change, and is developing new
Disaster Epidemiology tools for MDH-HPCD.
Mary Jeanne Levitt, MBC, is the communications coordinator with MN Tracking. She has a
Master’s in Business Communications and has worked for over 20 years in both the public and
non-profit sector in project management of research and training grants, communications and
marketing strategies, focus groups and evaluations of educational needs of public health
professionals. She serves on three institutional review boards, which specialize in academic
research, oncology research, and overall clinical research.
Paula Lindgren, MS, received her Master’s of Science Degree in Biostatistics from the University
of Minnesota. She works for the Minnesota Department of Health as a biostatistician and
provides statistical and technical support to MN Tracking for data reports, publications, webbased portal dissemination and presentations in the Chronic Disease and Environmental
Epidemiology section. Ms. Lindgren has also received training in the area of GIS for chronic
disease mapping and analysis. In addition to her work for MN Tracking, she works for various
programs within Chronic Disease and Environmental Epidemiology including the Asthma
program, Center for Occupation Health and Safety, Minnesota Cancer Surveillance System and
Cancer Control section.
Matthew Montesano, MPH, the Data Portal Coordinator with the Minnesota Tracking Program,
is responsible for the Data Portal’s content strategy, ensuring that its utility is maximized
through evidence-based health and science communications practices. He has expertise in
communicating health and science to lay audiences and developing strategic web-based public
health material. He is an advocate for the use of plain language and data visualization
techniques that increase users’ understanding of complex information. He has over 8 years of
nonprofit and public health experience with community programming, research, and
evaluation.
Jessica Nelson, PhD, is an epidemiologist with MN Tracking, working primarily on design,
coordination, and analysis of biomonitoring projects. Jessica received her PhD and MPH in
Environmental Health from the Boston University School of Public Health where her research
involved the epidemiologic analysis of biomonitoring data on perfluorochemicals. Jessica was
the coordinator of the Boston Consensus Conference on Biomonitoring, a project that gathered
input and recommendations on the practice and uses of biomonitoring from a group of Bostonarea lay people.
Christina Rosebush, MPH, is an epidemiologist with MN Tracking. Her work includes the
development and coordination of biomonitoring projects that assess perfluorochemicals (PFCs)
and mercury in Minnesota communities. She also works on collection and statistical analysis of
public health surveillance data for MN Tracking, with a focus on behavioral risk factors.
Christina received her Master’s Degree in Epidemiology from the University of Minnesota’s
School of Public Health, completing research in PFC biomonitoring for the Minnesota
Department of Health in partial fulfillment of her degree.
58
Jeannette M. Sample, MPH, is an epidemiologist with MN Tracking at the Minnesota
Department of Health, working primarily with the collection and statistical analysis of public
health surveillance data for MN Tracking. She also works on research collaborations with
academic partners relating to reproductive outcomes and birth defects. Prior to joining MN
Tracking, she was a CSTE/CDC Applied Epidemiology Fellow with the MDH Birth Defect
Information System. Jeannette received her Master’s Degree in Epidemiology and Biostatistics
from The George Washington University in Washington, DC.
Blair Sevcik, MPH, is an epidemiologist with MN Tracking at the Minnesota Department of
Health, where she works on the collection and statistical analysis of public health surveillance
data for .MN Tracking. Prior to joining MN Tracking in January 2009, she was a student worker
with the MDH Asthma Program. She received her Master of Public Health Degree in
epidemiology from University of Minnesota School of Public Health in December 2010.
Chuck Stroebel, MSPH, is the MN Tracking Program Manager. He provides day-to-day direction
for program activities, including: 1) development and implementation of the state network,
2) development and transport of NCDMs and metadata for the national network, and
3) collaboration and communication with key EPHT partners and stakeholders. Chuck received
a Masters of Public Health in Environmental Health Sciences from the University of North
Carolina (Chapel Hill). He has over 15 years of expertise in environmental health, including
areas of air quality, pesticides, climate change, risk assessment, and toxicology. Chuck also
played a key role in early initiatives to build tracking capacity at the Minnesota Department of
Health. Currently, he is a member of the IBIS Steering Committee (state network), the MDH
ASTHO Grant Steering Committee (climate change), and the Northland Society of Toxicology. He
also serves on the Minnesota EPHT Technical and Communications Teams.
Janis Taramelli, TTS, is the Community Outreach Coordinator for MN Biomonitoring, responsible
for communications with the MN Tracking Advisory Panel and study participants. A tobacco
treatment specialist, she has 20 years of experience working on research studies, surveys,
group facilitation, and one-on-one counseling in both the public and private sectors. Her
background includes development and coordination of statewide QUITPLAN at Work programs,
metro area QUITPLAN centers, and piloting tobacco cessation and heart healthy programs for
Minnesota’s Sage (Breast and Cervical Cancer Screening) and SagePlus (Heart Health Screening)
programs, funded by the Centers for Disease Control.
Addis Teshome has been an epidemiologist with MN Tracking since September 2014. Her work
involves populating a database of the scientific literature on perflurochemicals (PFCs),
performing statistical analysis of public health data, and developing various elements of the MN
Family Environmental Exposure Tracking project. Prior to joining MN Tracking as a student
worker in June 2014, she held similar positions at MDH’s Center for Occupational Health and on
the Safety and the Autism Spectrum Disorders Public Health Surveillance Report. Addis is
analyzing trends in predictors and outcomes of alcohol consumption among racial/ethnic
subgroups in partial fulfilment of her master’s degree in epidemiology at the University of
Minnesota’s School of Public Health.
Allan N. Williams, MPH, PhD, is an environmental and occupational epidemiologist in the
Chronic Disease and Environmental Epidemiology Section at the Minnesota Department of
59
Health. He is the supervisor for the MDH Center for Occupational Health and Safety. For over 25
years, he has worked on issues relating to environmental and occupational cancer, cancer
clusters, work-related respiratory diseases, and the surveillance and prevention of work-related
injuries among adolescents. He has served as the PI on two NIOSH R01 grants, as a coinvestigator on four other federally-funded studies in environmental or occupational health,
and is an adjunct faculty member in the University of Minnesota’s School of Public Health. He
received an MA in Biology from Indiana University, an MPH in Environmental Health and
Epidemiology from the University of Minnesota, and a PhD in Environmental and Occupational
Health from the University of Minnesota.
60