Does Synthetic Turf Rubber Crumb Cause Cancer?
Incidence of Malignant Lymphoma in Adolescents and Young Adults (AYAs)
in 18 States and Regions of the United States with Varying Synthetic Turf Field Density
and Selected Counties in California with Highest Densities
Archie Bleyer, MD
Clinical Research Professor
Oregon Health and Science University
Professor of Pediatrics
University of Texas Medical School at Houston
Past Chair, Children’s Cancer Group
World’s largest pediatric cancer research organization
Theresa Keegan, PhD, MS
Associate Professor, Department of Pediatrics
University of California Davis Comprehensive Cancer Center
June 5, 2016
(modified November 24, 2016)
Background
An associate coach at the University of Washington has collected cases of Hodgkin and non-Hodgkin
lymphoma in adolescent and young adult (AYA) soccer players who practiced and played on
synthetic turf fields, some of whom were goalkeepers. Thus synthetic turf and particularly the crum
rubber derived from tires in the material, and to which
Figure 1
goalkeepers are more exposed,
Prevalence of the Suspect Cancers in AYA Soccer Players
is thereby suspected
of causing lymphoma.
Lymphoma is the
most common
malignancy in the
age group (Fig. 1)
and hence it would
be expected to be
the cancer in the age
group. The question
is whether the
incidence is higher
than expected and
thus can implicate
the fields.
Null-Hypothesis
Regions in the U.S. with a higher prevalence of synthetic turf fields have had an
increasing and higher incidence of lymphoma among adolescents and young adults.
Methods
• Because lymphoma is strongly race/ethnicity dependent (Fig. 2), variation of
lymphoma by environmental exposure has to be evaluated by race/ethnicity.
• The 18 National Cancer Institute SEER registries are distributed throughout the U.S. to
enable race/ethnicity analysis and should therefore be evaluable for the prevalence of
synthetic turf fields, given the regions with the highest socioeconomic status
(California metropolitan areas, Seattle/Puget Sound) and adverse weather (SeattlePuget Sound) have the greatest prevalence (e.g., Los Angeles and San Francisco Bay
Area have 856 fields) whereas the southern and rural states (e.g., Louisiana, Georgia,
New Mexico, Utah, Hawaii) have the lowest incidence (Appendix 3). Connecticut and
metropolitan Atlanta were analyzed as either having high, intermediate or low
prevalence since these areas were more difficult to predict for synthetic field density
than other regions.
• Both the county-level incidence of cancer since 2000 and current number of synthetic
turf fields (Supplementary Appendix 1 ) are known in California, as well as the race
and ethnicity of persons living in the state, thereby enabling all 58 California counties
to be assessed for a relationship between lymphoma incidence and synthetic turf field
density (number of fields per population).
Methods (continued)
• The incidence of lymphoma in AYAs is also directly proportional to the socioeconomic
status of the community in which the AYA lived (Fig. 3).
• Hence the 18 SEER regions were evaluated for lymphoma incidence since 2000
among AYA age groups by race/ethnicity, socioeconomic status, and region.
Figure 2
Incidence of Lymphoma (Hodgkin and Non-Hodgkin) in Middle School, High School and
College Age Persons, U.S. SEER, 2000-2013, by 5-Year Age Intervals and Race/Ethnicity
10-14 years
Age at
Diagnosis
Non-Hispanic White
Black
Hispanic
Asian
North Amer Native
15-19 years
20-24 years
25-29 years
0
1
2
3
4
5
Incidence per 100,000 per Year
6
Figure 2 (continued)
Incidence of Malignant Lymphoma during 2000-2013 in 14- to 30-Year-Olds, U.S. SEER,*
by Race/Ethnicity
Age 14-30 Years
Incidence of Malignant Lymphoma in 14- to 30-Year-Olds, by Race/Ethnicity
Non-Hispanic White
Black
Race/Ethnicity
Hispanic
Asian
Native North American
0
1
2
3
4
5
6
7
8
Incidence per 100,000, Age-Corrected
Figure 3
Incidence of Malignant Lymphoma during 2000-2013 in 14- to 30-Year-Olds, U.S. SEER,*
by Race/Ethnicity and Cost-of-Living-Adjusted Median Family Income, 2000
*18 Regions
Non-Hispanic White
Race/Ethnicity
High 6955-9723
Intermediate 4187-6954
Low 1417-4186
Hispanic
Black
Asian
0
2
4
6
Incidence per 100,000, Age-Corrected
8
10
12
U.S. SEER, 2000-2013
Figure 4 depicts the incidence of lymphoma in 14- to 30-year-olds across 18 regions of the U.S.
according to the estimated local synthetic turf field density for each of the major race/ethnicities
in the U.S. There is no correlation between the estimated synthetic turf density and the incidence
of lymphoma in AYAs who are most likely to use such fields.
Figure 4
Incidence of Malignant Lymphoma (Hodgkin and Non-Hodgkin) in High School and College Age
Persons (Age 14-30), by Estimated Synthetic Turf Field Prevalence, U.S. SEER,* 2000-2013
High
Intermediate
Low
Estimated Synthetic Turf Field Prevalence:**
Non-Hispanic White
*18 Regions
7.2
3.9
Hispanic
3.9
Black
4.2
American Indian/Alaska
Native
1.2
1.2
0
1
7.5
4.3
Areas with more
Areas with more
synthetic turf
synthetic turf
fields do not
fields do not have
have a higher
a higher incidence
incidence of
of lymphoma
lymphoma
5.1
5.3
5.1
Asian or Pacific Islander
7.5
4.5
4.5
1.8
2
3
4
5
6
Incidence of Lymphoma per 100,000 per Year
7
8
**Assumptions based on socioeconomic status and weather (Appendix 2)
Figure 5
Incidence of Malignant Lymphoma in 14- to 30-Year-Oids, 2000-2013, SEER18,
by Estimated Synthetic Turf Field Prevalence, Race/Ethnicity, and Median Family Income
HIGH MEDIAN HOUSEHOLD INCOME (6000-9723 Rating)
8.5
Non-Hispanic White
When household income
within race/ethnicity groups is
applied to the SEER registries
grouped according to
likelihood of synthetic turf
field prevalence, there is also
no correlation between
lymphoma incidence and
synthetic turf field prevalence
(Fig. 5).
9.0
8.2
5.3
Hispanic
Areas with more
synthetic turf
fields do not have
a higher incidence
of lymphoma
4.5
4.9
6.6
Black
7.1
6.9
5.1
4.8
Asian
Estimated Synthetic
Turf Field Prevalence
4.0
High
Intermediate
Low
LOW MEDIAN HOUSEHOLD INCOME (1417-5999 Rating)
7.6
Non-Hispanic White
The lack of correlation was
apparent whether Connecticut
and Atlanta were considered as
high, intermediate or low
synthetic-field prevalence areas.
7.2
6.3
4.2
4.4
Hispanic
3.8
6.6
Black
6.2
5.6
Areas with more
synthetic turf fields do
not have a consistently
higher incidence of
lymphoma
3.4
Asian
3.8
4.6
0
1
2
3
4
5
6
7
8
Incidence of Lymphoma per 100,000 per Year
9
Annual Incidence of Malignant Lymphoma among 14 to 30 Year-Olds
in California’s Counties with the Greatest Density of Synthetic Turf Fields,
1974-2013 (SEER9) and 1992-2013 (SEER13), All Races/Ethnicities
Synthetic turf fields with crumb rubber were first installed in 1997 and more than 900
have since been installed in California. If they are strong lymphomagenic vectors, counties
with most of the fields and highest usage should have had some evidence for a trend since
1997 in an increase in lymphoma incidence in adolescents and young adults (AYAs).
• As shown by the brown SEER9 data in Figure 6, there has been no evidence for an
increase in lymphoma incidence among AYAs in the two California counties with the
highest density of synthetic turf fields, San Mateo and Marin (56 fields in a 2-county area
of 190,000 14-30 year-olds), during the 40 years form 1974 to 2013.
• As shown by the purple SEER13 data in Figure 6, there is no evidence for an increasing
incidence trend since 1992 in AYAs in the 8 counties with the highest density of synthetic
turf fields (219 fields in a population of 970,000 14-30 year-olds) .
• Compared to AYA females, AYA males are more likely to have had an increase since they
have a greater exposure to the fields. In this ecologic analysis they also have a greater
sensitivity to demonstrate a turf field effect. As shown by the blue data in Figures 7 and
9, neither the 2 highest density counties nor the 8 counties among the highest tier of
densities have evidence for an increase in lymphoma since 1992 in males (or females).
These ecologic findings are inconsistent with a hypothesis that synthetic turf fields
containing crumb rubber cause lymphoma in AYAs.
California, 1975-2013
Figure 6
Annual Incidence of Malignant Lymphoma among 14 to 30 Year-Olds in California’s Counties with the
Greatest Density of Synthetic Turf Fields, 1974-2013 (SEER9) and 1992-2013 (SEER13),
All Races/Ethnicities
Incidence of Malignant Lymphoma per 100,000 Population per Year
3-year moving averages
Dashed lines: linear egressions
12
Neither regression
slopes upward at a statistically
significant rate
10
8
San Mateo and Marin
Counties
APC p-Value
6
0.42
2000-2013 -0.25
0.45
San Mateo, Marin, Contra Costa, Shasta, Plumas,
Mono, Santa Clara, and Monterey Counties
4
Installation of Synthetic Turf Fields
with Crumb Rubber
2
0
1975
1974-2013 1.6
1980
1985
1990
1995
2000
2005
2010
Figure 7
Annual Incidence of Malignant Lymphoma among 14 to 30 Year-Olds in the Two California
Counties with the Greatest Density of Synthetic Turf Fields, San Mateo and Marin Counties,
during 1974-2013 (SEER9), All Races/Ethnicities
Incidence of Malignant Lymphoma per 100,000 Population per Year
3-year moving averages
Dashed lines: linear regressions
14
Neither regression
slopes upward at a statistically
significant rate
12
Males
10
APC p-Value
-0.40 0.76
8
1.85
6
4
Females
2
0
1974
Installation of Synthetic Turf Fields
with Crumb Rubber
1978
1982
1986
1990
1994
1998
2002
2006
2010
0.25
Figure 8
Annual Race and Ethnicity Population Distribution of among 14 to 30 Year-Olds in California’s Two
Counties with the Greatest Density of Synthetic Turf Fields, 1974-2013
San Mateo and Marin Counties
Other
Black
White
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
Asian
Black
Hispanic
Non-Hispanic White
Females
Hispanic
Non-Hispanic White
2012
2010
2008
2006
2004
2002
Installation of
Synthetic Turf Fields
with Crumb Rubber
2000
White
Black
1998
Black
Asian
1996
Other
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
1994
Females
Installation of Synthetic Turf
Fields with Crumb Rubber
Males
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
1992
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
Males
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
Figure 9
Annual Incidence of Malignant Lymphoma among 14 to 30 Year-Olds in California’s SEER13 Counties
with the Greatest Density of Synthetic Turf Fields, 1992-2013, All Races/Ethnicities
San Mateo, Marin, Contra Costa, Santa Clara and Monterey Counties
2-year moving averages Dashed lines: linear regressions
Incidence of Malignant Lymphoma per 100,000 Population per Year
12
Neither regression
slopes upward
10
Males
8
APC p-Value
`
6
-0.25 0.47
-0.21 0.68
Females
4
2
Installation of Synthetic Turf Fields with Crumb Rubber
0
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
Figure 10
Annual Race and Ethnicity Population Distribution of among 14 to 30 Year-Olds in California’s SEER13
Counties with the Greatest Density of Synthetic Turf Fields, 1992-2013
San Mateo, Marin, Contra Costa, Santa Clara and Monterey Counties
100%
Asian
90%
80%
70%
Males
Black
60%
50%
Hispanic
40%
30%
Non-Hispanic White
20%
0%
100%
90%
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
10%
80%
Black
70%
60%
Hispanic
50%
40%
30%
20%
10%
Installation of Synthetic Turf Fields
with Crumb Rubber
0%
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
Females
Asian
Non-Hispanic White
Figures 8 and 10 show the racial/ethnic population distribution in the two and five
counties with the greatest synthetic turf field densities as a function of calendar year
since 1974 for whites, blacks and other races and since 1992 for non-Hispanic
whites, Hispanics, Asians and black when the race/ethnicity data became available.
In general the changes in race/ethnicity distribution do not affect the overall finding
of a lack of correlation between the development of synthetic turf fields containing
crumb rubber and the incidence of malignant lymphoma in California’s counties with
the highest density of fields. There was a trend in a decrease in the non-Hispanic
white proportion that in the presence of no change in lymphoma incidence could be
argued to be consistent with lymphomagenic hypothesis, but the decrease in the
non-Hispanic white proportion from 1997, when the suspect fields were first
installed to 2009 was 7-9% (Figs. 8 and 10), ceased after 2009 (Figs. 8 and 10), and in
the two counties with the greatest field densities, increased thereafter (Fig. 8).
Discussion
The lack of correlation of lymphoma incidence with estimated synthetic turf field
density across the United States in the 18 states and regional areas monitored
closely by the National Cancer Institute with the prevalence of such fields in any of
the major races and socioeconomic categories is evidence against a causal
relationship. The lack of incidence trend consistent with the implementation of the
fields in the counties in California with the greatest density of synthetic turf fields
supports the lack of correlation in the national data.
Synthetic turf fields are more likely in metropolitan areas, in regions of higher
socioeconomic status that can afford such fields, and where local weather favors
synthetic fields. Figures 2 and 3 predict that lymphoma should be more frequent in
adolescents and young adults who live near synthetic turf fields because of their
race/ethnicity and socioeconomic status. Thus, middle-school, high-school and
college age soccer players who live in areas with synthetic turf fields would be
expected to have a higher rate of malignant lymphoma because of the
race/ethnicity and socioeconomic status and not because of the fields. The
incidence of malignant lymphoma in this age group of soccer players is consistent
with this relationship, as shown in Figures 4-8 , and thus mitigates against synthetic
turf is a causative factor.
Conclusion
The null-hypothesis is rejected: The incidence of malignant lymphoma in older
adolescents and young adults is not increased in regions with a higher prevalence
of synthetic turf fields. If increased, it is due to the race, ethnicity and
socioeconomic status of those who live in areas that have such fields and likely
not to the fields. This population level study does not eliminate crumb rubber in
synthetic turf as a potential contributing factor to the development of lymphoma
in those who use the fields, but it is consistent with a lack of such effect.
Limitations of This Type of Ecologic Analysis
As to how sensitive this ecologic type of analysis may be, the counties with the
highest synthetic turf field densities allow some estimates, based on the following
assumptions and deductions.*
• A soccer game involves 25-28 players; multiple games are played each weekend; a
practice session averages 40 persons; at least 2 practice sessions are held each
weekday; there are 3 seasons (Spring, Fall, Summer) of different leagues of players.
• If the players use the fields twice weekly (1 practice and 1 game), and each field
averages 10 teams per week, and there are three league seasons, the number of
exposed players is approximately 33,600 (56 fields x 20 players per team x 10 teams
per week x 3 seasons/leagues).
• That’s 18% of the 190,000 14-30 year-olds in the two counties, San Mateo and
Marin counties with 56 fields.
• For the five counties in the highest tier density with 219 fields, the comparable
values is 14% of 970,000 AYAs.
• At 14-18%, the ecologic findings should be able to detect a strong lymphomagenic
effect of synthetic turf fields in the evaluated counties.