[ExI] Paper: Citizen Science Genomics as a Model for Crowdsourced Preventive Medicine Research

[Bryan Bishop]

Citizen Science Genomics as a Model for Crowdsourced Preventive
Medicine Research
http://www.jopm.org/evidence/research/2010/12/23/citizen-science-genomics-as-a-model-for-crowdsourced-preventive-medicine-research/

Melanie Swan, Kristina Hathaway, Chris Hogg, Raymond McCauley & Aaron
Vollrath | Research | Vol. 2, 2010 | December 23, 2010

"""
Abstract

Summary: A research model for the conduct of citizen science genomics
is described in which personal genomic data is integrated with
physical biomarker data to study the impact of various interventions
on a predefined endpoint. This research model can be used for
large-scale preventive medicine studies by both institutional
researchers and citizen science groups. The genome-phenotype-outcome
methodology comprises seven steps: 1) identifying an area of
genotype/phenotype linkage for study, 2) conducting a thorough
literature review of data supporting this genotype/phenotype linkage,
3) elucidating the underlying biological mechanism, 4) reviewing
related studies and clinical trials, 5) designing the study protocol,
6) testing the study design and protocol in a small pilot study, and
7) modifying study design and protocol based on information from the
pilot study for a large-scale prospective study. This paper describes
a real-world example of the methodology implemented for a proposed
study of polymorphisms in the MTHFR gene, and how these polymorphisms
may influence homocysteine levels and vitamin B deficiency. The
current study looks at the possibility of optimizing personalized
interventions per the genotype-phenotype profiles of individuals, and
tests the hypothesis that simple interventions may be effective in
reducing homocysteine in individuals with high baseline levels,
particularly in the presence of a polymorphism in the MTHFR variant
rs1801133.

Keywords: MTHFR, homocysteine, genomics, polymorphism, variant,
citizen science, patient-driven clinical trial, crowdsourced clinical
trial, research study, self-experimentation, intervention,
personalized medicine, preventive medicine, participatory medicine,
quantified self, genome-phenotype-outcome study, citizen science
genomics.

Citation: Swan M, Hathaway K, Hogg C, McCauley R, Vollrath A. Citizen
science genomics as a model for crowdsourced preventive medicine
research. J Participat Med. 2010 Dec 23; 2:e20.
Published: December 23, 2010.
Competing Interests: The authors have declared that no competing
interests exist.

Introduction

Continually decreasing costs in genomic sequencing have made it
possible for individuals to obtain their own genomic data. An
estimated 80,000 individuals have subscribed to consumer genomic
services. Genotyping provider 23andMe counted 50,000 subscribers as of
June 2010.[1] Navigenics and deCODEme had an estimated 20,000 and
10,000, respectively, as of March 2010.[2] Others may be clients of
Pathway Genomics or other services. Today, individuals can view the
200 or so variants analyzed by consumer genomic companies for a
variety of disease, drug response, trait, and carrier status
conditions via a web-based interface, and a question naturally arises
as to what else can be done with the data.

Tools do not yet exist to identify and prevent disease before clinical
onset. Integration of genomic, phenotypic, environmental, and
microbiometric health data streams will be required to create reliable
predictive tools. The potential volume of this data is staggering,
numbering, perhaps, a billion data points per person,[3] which may
routinely generate zettabytes of medical data.[4]

The combination of multiple health data streams, the anticipated data
deluge, and the challenges and expense of recruiting subjects for
studies all suggest that there could be a benefit to supplementing
traditional randomized clinical trials with other techniques.[5]
Crowdsourced cohorts of citizen scientists (eg, patient registries)
could be a significant resource for testing multiple hypotheses as
research could be quickly and dynamically applied in various
populations. Engaged citizen scientists could collect, synthesize,
review, and analyze data. They could interpret algorithms, and run
bioinformatic experiments. This paper proposes a research model that
could be used in conducting citizen science genomics, that integrates
personal genomic data with physical biomarker data and interventions,
and that could be applied in large-scale preventive medicine studies
by both institutional researchers and citizen science groups.

Methods

An increasing number of individuals have access to their own genomic
data, would like to contribute this data to scientific research, and
would like to put it to use in managing their own health. Scalable
models for conducting citizen science studies are needed. The authors
designed a methodology for the conduct of citizen science genomics
which links genomic data to corresponding phenotypic measures and
relevant interventions. The purpose is to create mechanisms for
establishing and monitoring baseline measures of wellness, and tools
for the conduct of preventive medicine. The key steps in the
methodology include:

1. Selecting a specific area of genotype/phenotype linkage for
potential study and generating a testable hypothesis
2. Conducting a literature review to validate the selected study area
3. Analyzing the underlying biological pathway and mechanism
4. Reviewing related studies and clinical trials
5. Designing the study protocol
6. Testing the study design in a small non-statistically significant pilot
7. Identifying the next steps for a full-scale launch of the study

Results

The results are presented as a detailed outline of the seven-step
methodology for operating citizen science genomic studies. The
methodology is implemented in the specific case of a proposed study
looking at polymorphisms in the MTHFR gene and how these polymorphisms
relate to homocysteine levels and vitamin B deficiency.

1. Select a specific area of genotype/phenotype linkage for potential
study and generate a testable hypothesis.

For the inaugural citizen science genomic study, 40 potential ideas
were identified in a variety of health and behavioral genomic areas in
recently published research (http://diygenomics.pbworks.com). One area
that seemed conducive to study was the potential association of the
MTHFR gene and vitamin B deficiency. MTHFR polymorphisms may keep
vitamin B-9 (folic acid) from being metabolized into its active form,
folate. This may lead to the potentially harmful accumulation of
homocysteine. There is a strong research-supported association between
the principal MTHFR variant (rs1801133) and homocysteine levels.[6]
Genotyping data for MTHFR variants are available in 23andMe data.
Furthermore, blood tests for homocysteine, vitamin B-12, and vitamin
B-9 are readily obtainable, as are over-the-counter vitamin supplement
interventions. A testable hypothesis was generated that supplements
may be effective in reducing homocysteine levels, particularly for
those with a genetic polymorphism.

Studying MTHFR and vitamin B deficiency could have an important public
health benefit since approximately half of the US population is
estimated to have one or more MTHFR polymorphisms. The distribution of
genotypes in the US for rs1801133 is 49% CC (homozygous normal), 40%
CT (heterozygous), and 11% TT (homozygous risk).[7] In addition,
vitamin B-12 deficiency is a common nutritional deficiency in both the
US and the developing world,[8] particularly for the elderly and
vegetarians (approximately 3% of the US population).[9]

2. Conduct a literature review to validate the selected study area

Numerous observational and prospective studies have found correlations
between elevated plasma homocysteine levels and cardiovascular
disease, renal disease, depression, anxiety, Alzheimer’s disease, and
colorectal cancer.[10][11][12][13][14]

The majority of published literature relates to cardiovascular
disease. A meta-analysis of 30 prospective and retrospective studies
(involving a total of 5,073 ischemic heart disease (IHD) events and
1,113 stroke events) showed that a 25% lower homocysteine level was
independently associated with an 11% lower risk of coronary heart
disease and a 19% lower risk of stroke.[10] Despite this, the causal
relationship between elevated homocysteine and cardiovascular outcomes
has not been conclusively proven. A large (n=12,064), recently
published (June 2010), prospective, randomized study of patients with
a prior myocardial infarction provided either folic acid or vitamin
B-12 supplementation compared to placebo. The authors tracked coronary
events over an average of 6.7 years. They found an average reduction
of 28% in plasma homocysteine levels, but no difference between the
vitamin group and placebo group in the occurrence of coronary events
or death.[15] However a prospective, randomized study of the impact of
homocysteine levels on the progression of atherosclerosis showed that
folic acid supplementation led to reduced homocysteine levels and a
regression in carotid intima-media thickness (CIMT) compared to an
increase in CIMT for the placebo group.[16]

Although more research is needed, there appears to be adequate
evidence that low homocysteine levels are desirable, and may reduce
risk for a number of conditions.

3. Analyze the underlying biological pathway and mechanism

The MTHFR pathway and homocysteine metabolism are the underlying
biological mechanisms in this study. There are a number of ways in
which genetic variation and intervention may impact homocysteine
metabolism. Homocysteine is a naturally-occurring amino acid in the
blood which is broken down (metabolized) through three interconnected
pathways: the folate cycle, methionine cycle, and transsulfuration
pathway (Figure 1).[17] A detailed explanation of homocysteine
metabolism is presented in the Supplementary Material. The pathways
are fairly complex and involve two other enzymes in addition to MTHFR.
It is possible that different interventions could impact overall
homocysteine metabolism in different ways. In Figure 1, the red boxes
show the different places where the first intervention (the inactive
form of B-9, further described below) may impact the pathway; the
green box shows where the second intervention (the active form of B-9)
may impact the pathway.

Figure 1: Homocysteine metabolism.

4. Review related studies and clinical trials

Several clinical trials have been conducted to investigate the ability
of interventions to lower homocysteine levels. A detailed review of
nine studies was conducted and is presented in the Supplementary
Material. The average overall result was to lower homocysteine by 23%.
Two studies[18][19] specifically compared folic acid with the active
form of folate, 5-MTHF (5-methyltetrahydrofolate). Both found that the
active formulation was most effective in reducing homocysteine levels
(Akoglu 37% versus 24%;[18] Lamers 19% versus 12%[19]).

The existing clinical trials suggest that several factors may
influence baseline homocysteine levels, in particular, age, health
status, and genotype. Individuals who were older (especially over 50),
had just experienced a major health disruption, or had one or more
polymorphisms in the main MTHFR variant rs1801133, were more likely to
have higher baseline homocysteine levels than those that did not
(Supplementary Material: Figure 2). Further, the reduction proportion
from the baseline level was greater for those individuals with higher
initial levels of homocysteine.
5. Design the study protocol

The required genomic and phenotypic data were identified.
Approximately 20 variants have been linked with homocysteine in
genome-wide association studies.[20] MTHFR 677C>T (rs1801133) was
selected as the variant with the strongest association to mild enzyme
deficiency, and MTHFR 1298A>C (rs1801131) as the leading secondary
variant.[6] The corresponding phenotypic measures selected were blood
tests for homocysteine, vitamin B-12, and folate (vitamin B-9).

The type and timing of interventions were determined based on
published literature. The background research on the MTHFR mechanism
suggests that individuals with one or more polymorphisms may not be
able to metabolize folic acid (the inactive form of B-9) into its
active form (tetrahydrofolate or folate), as efficiently as
individuals without a polymorphism. Therefore, the first intervention
selected was administration of the inactive form of B-9, which is
commonly present in over the counter B vitamin products such as
Centrum multivitamins. The second intervention involved administration
of the active form of folate (L-methylfolate), and the third was
administration of the inactive and active forms together (also being
tested by a current clinical trial).[21] The supplement contents were
as follows: the Centrum multivitamin contained 2 mg of pyridoxine
hydrochloride (B-6), 400 mcg of folic acid (B-9), and 6 mcg of
methylcobalamin (B-12); the Life Extension Foundation L-methylfolate
contained 1,000 mcg of L-methylfolate. The interventions were to be
taken on a daily basis, at the same time of day, with food.

For this pilot phase of the study, the authors opted to use a
crossover study design. Each individual tried each intervention, in
sequence, essentially serving as his or her own control. While other
homocysteine clinical trials typically had at least four-week periods
for testing interventions, two representative trials confirmed that
most of the observed effect occurred within the first two
weeks.[18][19] Therefore in the pilot study, two-week minimum
intervention periods were selected with a two-week washout period at
the beginning.

Participant recruitment was accomplished by talking about the study in
public speaking engagements and targeting special interest groups such
as the DIYbio, Quantified Self, Health 2.0, Singularity University,
futurist, and life extension communities, particularly 23andMe
clients. Some potential participants were motivated to sign up for
23andMe in order to participate in citizen science genomic studies.
Many potential participants were interested, but did not join the
study for a variety of reasons. The biggest barrier was the
self-supported cost of blood tests and supplement interventions
($291). In a full-scale launch, other strategies will be necessary to
target a more representative segment of the population.

6. Test the study design in a small non-statistically significant pilot

To test the study design, a small non-statistically significant pilot
study was conducted in three phases: execution, results collection,
and results analysis. The type of analysis that could be conducted on
data results is presented here, realizing that the pilot cohort sample
size (n=7) is not statistically significant.

Seven healthy men and women, ages 26-47, who had not taken any vitamin
supplements for two weeks or more and met other usual study exclusion
criteria, were enrolled in the study. The study was conducted from
June to December 2010. Three participants cycled through the study at
nearly exact two-week intervals. Three participants went through the
study in two- to three-week periods on average, and one participant
specifically tested three-week intervals. Six participants ordered
blood tests from the Life Extension Foundation as they offered the
lowest cost, and lab work orders were fulfilled at local LabCorp
(standardized testing) facilities in the US. The remaining participant
had homocysteine levels tested at a Japanese medical facility in
Tokyo. The L-methylfolate supplement was mail-ordered by the group
from the Life Extension Foundation. The Centrum multivitamin was
purchased individually at local drug stores. All seven of the study
participants collaborated in the study design or an active review of
the protocol.

The study relied on self-reporting that the supplement protocol was
followed. Participants tried to avoid unusual variance in nutrition,
exercise, stress levels, sleep, and other behaviors. Participants
looked up their genotype data for the relevant MTHFR variants in their
23andMe data files (genotyping is assumed to be accurate[22]), and
recorded them in the study’s public wiki
(http://diygenomics.pbworks.com/MTHFR_Results). Blood test
measurements from LabCorp PDFs or other reports were entered similarly
in the public wiki. All participants were interested in full
transparency and public accessibility of their genotypic and
phenotypic study results, and allowed their names to be associated
with the study. Participants were enumerated as Citizen 1, 2, etc with
their initials.

Genotype results: Table 1 lists the pilot study participants and their
genotype data for the two reviewed variants. For the main associated
variant, rs1801133, three participants are homozygous normal (GG) and
four are heterozygous (AG). Two of the heterozygous participants are
also vegetarians/vegans which further increases their potential risk
of vitamin B deficiency. For the secondary variant, rs1801131, two
participants are homozygous normal (TT), four are heterozygous (GT),
and one is homozygous for the polymorphism (GG). The table then
includes maternal and paternal haplotype group information from
23andMe and demographic information regarding participant ethnicity,
gender, age, and vegetarian status.

23andMe’s genotype reporting method (all genotypes are listed as their
forward strand values) means that sometimes their genotyping values
need to be mapped to other conventions for interpretation. Commonly
used resources for obtaining major/minor allele mappings indicate C/T
as the major/minor alleles for rs1801133, and A/C for rs1801131
(dbSNP;[23] SNPedia;[24] HuGE Navigator[7]). The mapping of the
alleles from the standard resources to 23andMe would be that rs1801133
C/T is G/A in 23andMe data, and rs1801131 A/C is T/G in 23andMe data
(C maps to G and vice versa; A maps to T and vice versa). The mapping
was confirmed by comparing deCODEme, Navigenics, and 23andMe data
files for the same individuals, and by reviewing genotype prevalence
across multiple 23andMe files.

Table 1: Genotype results and demographic profiles.

Phenotype results: Figure 2 and Table 2 illustrate how homocysteine
levels shifted during the pilot study. Table 3 contains the blood test
data for vitamin B-12. At baseline, homocysteine levels ranged from
6.4 – 14.1 µmol/L. The cohort mean was 10.4 (SD (standard deviation)
3.03), and was higher for vegan/vegetarian individuals with a
polymorphism in rs1801133 (12.8 versus 9.5). After the first
intervention (Centrum multivitamin), homocysteine went down for six
individuals and up for one individual, and had a tighter range
(5.7-10.6; mean 8.8; SD 1.50).

After the second intervention (L-methylfolate), homocysteine was
higher for five individuals, including the four with a polymorphism,
and lower for two (mean 10.3; SD 2.77). For the four individuals that
included a plasma folate test, levels were at or above the high point
of the test reference range (19.9 mg/mL) (Supplementary Material –
Table 3) after the second intervention. After the third intervention
(Centrum multivitamin + L-methylfolate), for three of the four
participants who tried it, homocysteine was higher than with
L-methylfolate alone. In the final step, five individuals completed an
ending washout blood test, with three participants, including two with
a polymorphism, having lower homocysteine levels than after the third
intervention. The fourth participant had slightly higher homocysteine,
and the fifth participant had markedly higher homocysteine as compared
with the last intervention tried, the L-methylfolate. For three out of
four participants that included the vitamin B-12 test (Table 3), B-12
levels went up an average of 17.5% after the first intervention, and
one participant’s went down 17%. B-12 movement then generally
progressed flat or with a slight increase for the duration of the
study.

An analysis of the test data results was performed to calculate the
percent declines for each period from baseline and for each period
relative to the prior period (smoothing was employed for one missing
value). There was a 19% average decline in homocysteine for the best
solution in any period versus the baseline (Table 2) and a 21% average
decline in homocysteine for the best solution in any period versus the
prior period. There was not a significant difference between
homozygous normal individuals (GG) for the main variant rs1801133 (18%
average reduction) and heterozygous individuals (AG) (19% reduction),
but the two vegan/vegetarian heterozygous individuals experienced a
28% average reduction. In a larger study that investigated genotype
polymorphisms, a difference was found in having greater reduction in
heterozygous subjects (12% versus 9%).[25] The secondary variant,
rs1801131, did not seem to have an impact, either in isolation or when
considered together with rs1801133.

Figure 2: Participant homocysteine levels at study intervals.

Table 2: Participant homocysteine levels (µmol/L) and statistics at
study intervals.

Table 3: Participant vitamin B-12 levels (pg/mL) at study intervals.

Discussion

The overall result seen in this was a 19% average reduction in
homocysteine levels. While not statistically significant, this is
consistent with the 23% average reduction achieved in reported
clinical trials.

While a homocysteine range of 0.0-15.0 µmol/L is considered clinically
normal, many scientists contend that lower levels are preferable.
Suggested preferred levels are less than 11.4 µmol/L for men and less
than 10.4 µmol/L for women in one paper cited.[26] According to these
measures, four out of the seven pilot participants had high baseline
homocysteine levels which they were able to meaningfully reduce with
supplement interventions.

The best intervention for five out of seven individuals was the
regular B vitamin as opposed to the active form of B-9 (folate). The
active form of B-9 worked better for one individual. The remaining
individual, having a homozygous minor variant form of rs1801133, did
not have high initial homocysteine and found that the active form of
B-9 was better than the regular B vitamin, but that no intervention
was best. This suggests that targeted solutions may be optimum for
groups of individuals with certain profiles.

The biggest question was why the blood test values for homocysteine
increased in five individuals (three with high baseline homocysteine;
four with a polymorphism) after B-9 when other clinical trials found
the active form of B-9 to be the superior intervention for lowering
homocysteine. Participant behavior was generally consistent, and the
reproducibility of testing results (within-person, between-person, and
in labs) was also confirmed (Supplementary Material: Variability in
homocysteine test results). Variation could have been introduced by a
number of factors including the natural variability in homocysteine
levels, variability in the active ingredient amounts in the
intervention supplements, carryover effects between interventions
(also evidenced by the lack of homocysteine levels returning to
baseline levels in the final washout cases), or other complexities
related to the homocysteine pathway.

This pilot study represents an example of how
genomic-phenotypic-outcome research can be conducted in the era of
personalized genetic data availability. It also illustrates the
potential importance of including genomics as a data element in
preventive medicine research, and illustrates the potential of using
motivated individuals in citizen science genomic studies. Several
participants also indicated the value of their experience and how it
translated into post-study behavioral changes (Supplementary Material:
Personal statements from study participants).

7. Identify the next steps for a full-scale launch of the study

There are a number of steps required for a full-scale cohort launch
including implementing an independent ethical review and informed
consent process, adjusting the study protocol, forming strategies for
study financing and representative population targeting, and creating
a data collection and analysis platform:

Independent Ethical Review and Informed Consent
Citizen science genomics is human subjects research and as such,
should have independent ethical review and oversight. There are at
least two independent review boards in the US which have indicated
their willingness to discuss the potential review of citizen science
studies: IRC, Independent Review Consulting, Inc., in Corte Madera, CA
(http://www.irb-irc.com) and WIRB, the Western Institutional Review
Board, in Olympia, WA (http://www.wirb.com). A related model of
consumer genomic research conducted by 23andMe[27] brought up a number
of ethical issues,[28] and ultimately IRC reviewed their study. As
citizen science models develop, oversight models could evolve to
include citizen ethicists, citizen review boards, health advisors
(analogous to financial advisors), and insurance mechanisms for
personal health experimentation communities. Informed consent would be
an obviously required process to include in any full-scale human
subjects research study.

Protocol Adjustment

The pilot study confirmed that the central point for investigation in
a full-size cohort is whether interventions can be optimized according
to the genotype-phenotype profiles of individuals. The pilot study
also suggested that a regular B vitamin may be most effective in
lowering homocysteine in individuals with high baseline homocysteine
levels, especially in the presence of one or more rs1801133
polymorphisms. A number of structural changes could be made to improve
scientific rigor in a broader launch, including participant blinding,
inclusion of a placebo arm, and standardized monitoring, testing, and
interventions.

Strategies for Funding and Representative Population Targeting

To date, citizen science genomics has relied on the study recruitment
pool being the limited number of individuals (approximately 100,000)
who have subscribed to personal genotyping services. These individuals
may not be representative of the population at large; the literature
characterizes direct-to-consumer genomic customers as early adopters
and self-driven information seekers.[29][30] For widespread public
health studies, it will be necessary to target a broad diversity of
participants across multiple dimensions including information-seeking
and action-taking propensity, ethnicity, and socioeconomic background.
To accomplish this, traditional recruitment techniques could be used
together with new patient-centered social media strategies.
Conclusion

This paper presents citizen science genomics, a research model
contemplated for large-scale execution of preventive medicine research
in crowdsourced cohorts. The model integrates personal genomic data
with physical biomarker data to study the impact of various
interventions on a predefined endpoint. Citizen science genomics could
allow both traditional researchers and citizen scientists to access
crowdsourced subjects who are ready to engage in research studies.
Citizen scientists could be important resources as they increasingly
have access to their health information, may be willing to contribute
their data to various studies, have the interest and motivation to
investigate conditions of personal relevance, and can leverage
crowdsourced labor for data collection, monitoring, synthesis, and
analysis, and new tool development.

Preventive medicine is a key public health challenge in the coming
decades. New models like citizen science genomics are needed to answer
important questions. Dropping prices and new technologies for
collecting data regarding microbiomes, proteomics, imaging, personal
tracking, and other information streams will increase the feasibility
of this approach. Preventive medicine has the potential to take on new
relevance and meaning through the use of citizen science genomic
studies, as crowdsourced participants establish baseline and ongoing
longitudinal measures for wellness, health maintenance, and customized
intervention.

References

   1. Goetz T. Sergey Brin’s search for a Parkinson’s cure. Wired.
June 22, 2010. Available at:
http://www.wired.com/magazine/2010/06/ff_sergeys_search/all/1.
Accessed September 20, 2010. ↩
   2. Pollack A. Consumers slow to embrace the age of genomics. New
York Times. March 19, 2010. Available at:
http://www.nytimes.com/2010/03/20/business/20consumergene.html.
Accessed September 20, 2010. ↩
   3. Hood L. Systems medicine, transformational technologies and the
emergence of proactive P4 medicine. Paper presented at: Personalized
Medicine World Conference; January 19-20, 2010; Mountain View, CA.
Available at: http://pmwc2010.com/program.php. Accessed September 20,
2010. ↩
   4. Enriquez J. As the future catches you. Paper presented at: 2nd
Annual Consumer Genetics Conference; June 2-4, 2010; Boston, MA.
Available at: http://www.consumergeneticsshow.com/uploads/2010_Early_Schedule.pdf.pdf.
Accessed September 20, 2010. ↩
   5. Hartwell L. The promise and progress of personalized medicine.
Paperresented at the Sandra Day O’Connor College of Law Personalized
Medicine Conference; March 8-9, 2010; Scottsdale, AZ. Available at:
http://online.law.asu.edu/events/Personalized_Medicine. Accessed
September 20, 2010. ↩
   6. Sibani S, Christensen B, O’Ferrall E, et al. Characterization of
six novel mutations in the methylenetetrahydrofolate reductase (MTHFR)
gene in patients with homocystinuria. Hum Mutat. 2000;15(3):280-7. ↩
   7. Yu W, Yesupriya A, Chang M, et al. Genotype Prevalence Catalog.
HuGE Navigator. Available at:
http://www.hugenavigator.net/HuGENavigator/raceDisplay.do?submissionID=57&variationID=57.
Accessed September 20, 2010. ↩
   8. Harvard Health Publications. Vitamin B12 deficiency:
vegetarians, elderly may not get enough vitamin B12, says the Harvard
Health Letter. Available at:
http://www.health.harvard.edu/press_releases/vitamin_b12_deficiency.
Accessed November 29, 2010. ↩
   9. Vegetarian Times. Vegetarian Times Study Shows 7.3 Million
Americans Are Vegetarians and an additional 22.8 Million Follow a
Vegetarian-Inclined Diet (Data collected by the Harris Interactive
Service Bureau; data analysis performed by RRC Associates Colorado).
2008. Available at:
http://www.vegetariantimes.com/features/archive_of_editorial/667.
Accessed November 29, 2010. ↩
  10. Homocysteine Studies Collaboration. Homocysteine and risk of
ischemic heart disease and stroke: a meta-analysis. JAMA.
2002;288(16):2015-2022. ↩
  11. Stanger O, Fowler B, Piertzik K, et al. Homocysteine, folate and
vitamin B12 in neuropsychiatric diseases: review and treatment
recommendations. Expert Rev Neurother. 2009 Sep;9(9):1393-412. ↩
  12. Williams K, Schalinske K. Homocysteine metabolism and its
relation to health and disease. Biofactors. 2010 Jan;36(1):19-24. ↩
  13. Hooshmand B, Solomon A, Kåreholt I, et al. Homocysteine and
holotranscobalamin and the risk of Alzheimer disease: a longitudinal
study. Neurology. 2010 Oct 19;75(16):1408-14. ↩
  14. Zhu Q, Jin Z, Yuan Y, et al. Impact of MTHFR gene C677T
polymorphism on Bcl-2 gene methylation and protein expression in
colorectal cancer. Scand J Gastroenterol. 2010 Dec 6. ↩
  15. Study of the Effectiveness of Additional Reductions in
Cholesterol and Homocysteine (SEARCH) Collaborative Group, Armitage
JM, Bowman L, et al. Effects of homocysteine lowering with folic acid
plus vitamin B12 vs. placebo on mortality and major morbidity in
myocardial infarction survivors: a randomized trial. JAMA. 2010 Jun
23;303(24):2486-94. ↩
  16. Ntaios G, Savopoulos C, Karamitsos D, et al. The effect of folic
acid supplementation on carotid intima-media thickness in patients
with cardiovascular risk: a randomized, placebo-controlled trial. Int
J Cardiol. 2010 Aug 6;143(1):16-9. ↩
  17. Scott J, Weir D. Folic acid, homocysteine and one-carbon
metabolism: a review of the essential biochemistry. J Cardiovasc Risk.
1998;5(4): 223-7. ↩
  18. Akoglu B, Schrott M, Bolouri H, et al. The folic acid metabolite
L-5-methyltetrahydrofolate effectively reduces total serum
homocysteine level in orthotopic liver transplant recipients: a
double-blind placebo-controlled study. Eur J Clin Nutr. 2008
Jun;62(6):796-801. Page 798, Table 3. ↩
  19. Lamers Y, Prinz-Langenohl R, Moser R, et al. Supplementation
with [6S]-5-methyltetrahydrofolate or folic acid equally reduces
plasma total homocysteine concentrations in healthy women. Am J Clin
Nutr. 2004 Mar;79(3):473-8. ↩
  20. Paré G, Chasman DI, Parker AN, et al. Novel associations of
CPS1, MUT, NOX4, and DPEP1 with plasma homocysteine in a healthy
population: a genome-wide evaluation of 13,974 participants in the
Women’s Genome Health Study. Circ Cardiovasc Genet. 2009
Apr;2(2):142-50. ↩
  21. Flugelman, M. Examining B12 Deficiency Associated With C677T
Mutation on MTHFR Gene in Terms of Commonness and Endothelial
Function. Clinical trial in progress: NCT00730574. Available at:
http://clinicaltrials.gov/ct2/show/NCT00730574. Accessed September 20,
2010. ↩
  22. Ng PC, Murray SS, Levy S, et al. An agenda for personalized
medicine. Nature. 2009;461:724 –726. ↩
  23. National Center for Biotechnology Information (NCBI). dbSNP.
Available at: http://www.ncbi.nlm.nih.gov/sites/entrez?db=snp&cmd=search&term=rs1801133
and http://www.ncbi.nlm.nih.gov/sites/entrez?db=snp&cmd=search&term=rs1801131.
Accessed September 20, 2010. ↩
  24. Cariaso M. SNPedia. Available at:
http://www.snpedia.com/index.php/Rs1801133 and
http://www.snpedia.com/index.php/Rs1801131. Accessed September 20,
2010. ↩
  25. Ashfield-Watt PA, Pullin CH, Whiting JM, et al.
Methylenetetrahydrofolate reductase 677C–>T genotype modulates
homocysteine responses to a folate-rich diet or a low-dose folic acid
supplement: a randomized controlled trial. Am J Clin Nutr. 2002
Jul;76(1):180-6. ↩
  26. Selhub J, Jacques PF, Rosenberg IH, et al. Serum total
homocysteine concentrations in the third National Health and Nutrition
Examination Survey (1991-1994): population reference ranges and
contribution of vitamin status to high serum concentrations. Ann
Intern Med. 1999 Sep 7;131(5):331-9. ↩
  27. Eriksson N, Macpherson JM, Tung JY, et al. Web-based,
participant-driven studies yield novel genetic associations for common
traits. PLoS Genet. 2010 Jun 24;6(6):e1000993.↩
  28. Gibson G, Copenhaver GP. Consent and internet-enabled human
genomics. PLoS Genet. 2010 Jun 24;6(6):e1000965.↩
  29. McGuire AL, Diaz CM, Wang T, et al. Social networkers’ attitudes
toward direct-to-consumer personal genome testing. Am J Bioeth.
2009;9:3-10.↩
  30. McGowan ML, Fishman JR, Lambrix MA. Personal genomics and
individual identities: motivations and moral imperatives of early
users. New Genetics and Society. 2010 Sep;29(3):261-290.↩

Acknowledgments

We would like to acknowledge Takashi Kido and William Reinhardt for
sharing their genotypic and phenotypic data, and many individuals who
shared their genetic data for research purposes including David Orban,
Geoffrey Shmigelsky, Eri Gentry, Todd Huffman, Fadi Bishara, Richard
Leis, Jr., Mark Even Jensen, Misha Angrist, and several parties whom
wish to remain anonymous. We would like to acknowledge Lyn Powell and
Lucymarie Mantese for their advisory contribution and study support.

Copyright: © 2010 Melanie Swan, Kristina Hathaway, Chris Hogg, Raymond
McCauley, and Aaron Vollrath. Published here under license by The
Journal of Participatory Medicine. Copyright for this article is
retained by the author(s), with first publication rights granted to
the Journal of Participatory Medicine. All journal content, except
where otherwise noted, is licensed under a Creative Commons
Attribution 3.0 License. By virtue of their appearance in this
open-access journal, articles are free to use, with proper
attribution, in educational and other non-commercial settings.
"""

- Bryan
http://heybryan.org/
1 512 203 0507