What is needed for successful research in genetic epidemiology
What is needed for successful research in genetic epidemiology
Basic tools to perform genetic epidemiology are detailed information about the disease
phenotype, and biological samples for molecular research. However, these minimum requirements
might be not sufficient for successful results. Genetic epidemiology mainly deals
with complex diseases. On the one hand, most of them are relatively frequent in the
population. On the other hand, they are polygenic in origin, with several genes involved,
in addition to environmental influences like lifestyle and contact to toxic or carcinogenic
substances. Typically, the contribution of single genes as well as of single external
risk factors is small to moderate. Therefore, in addition to large sample numbers
a high quality of data as well as sound epidemiological methodology are prerequisites
of successful research. This is independent of the study type used (e. g. case control
study, case only study, cohort study, family-based linkage or association studies).
To study the genetic basis of complex diseases, typically thousands rather than hundreds
of patients with the disease of interest are needed. Therefore, the most common way
is to recruit them via hospitals or doctor’s offices. However, the quality of phenotypic
characterization might be variable. It is relatively easy to recruit large numbers
if the standards for diagnostic criteria are low, but if well defined patients are
needed, the available numbers are much smaller. To find a strong genetic effect, a
more crude phenotypic information might be sufficient, but for weak effects, thorough
phenotypic characterization is important.
The simplest way of phenotyping for the disease of interest is to take doctors’ diagnoses
from the records of the patients. However, it is quite clear for many diseases that
under the umbrella of one diagnosis several distinct pathophysiological entities are
summed up, which might have a different genetic origin. Therefore it is crucial for
genetic epidemiology to be based on well defined, standardized phenotyping, which
then allows us to define subgroups of patients that have more specific clinical or
sub-clinical properties. This makes it necessary to use specific laboratory parameters
as well as refined diagnostic tools.
Another strong argument for the importance of sound phenotyping is the matter of intermediate
phenotypes. Many parameters show an early reaction of the body, and they are associated
with the disease after its manifestation. They do, however, not yet represent a status
of disease. These parameters like specific IgE, elevated cholesterol, elevated C-reactive
protein or obesity are often called intermediate phenotypes. The chance of success
might be higher if we try to identify genes that influence intermediate phenotypes
instead of trying to identify the genetic influence on an endpoint which may be influenced
by dozens of intermediate phenotypes and therefore hundreds of genes.
As long as genetic epidemiology is „only” interested in the identification of genes,
most environmental factors may be ignored. However, the situation is different if
one is interested in gene environment interaction. In complex diseases, it is likely
that a combination of genes predisposing for the disease and environmental factors
exacerbating the impact of these genes are jointly responsible for disease development
in populations. In addition, environmental factors which seem to have only a moderate
impact at the population level might have larger relative risks in subpopulations
with certain genetic predispositions. Classical epidemiology has always been dealing
with these „environmental” risk factors, but only today we are able to combine knowledge
on the genetic background with classical epidemiological research, and we have tools
to investigate the interaction of genes and the environment whose applications help
to understand diseases [1].
International development of biobanking
International development of biobanking
In many countries large biobanks with more than 50,000 participants are planned or
have already been established (planned numbers in brackets). The first national biobank
of this size was realized in Iceland (270,000; [2]), followed by the Estonian project (100,000; [3]). The planning of the UK Biobank (500,000, [4]) is advanced, the Japanese biobank (500,000; [5]) has been started, and now in the US a biobank also is discussed (500,000 or more;
[6]). As one example the UK biobank will be described in the following.
The aim of the UK Biobank is to investigate the separate and combined effects of genetic
and environmental factors (including lifestyle, physiological and environmental exposures)
on the risk of common multifactorial diseases of adult life. At least 500,000 men
and women aged 45 to 69 years from the general population of the United Kingdom will
be investigated prospectively. People registered with participating general practices
will be asked to join the study by completing a questionnaire and attending an interview
and examination by a research nurse. Moreover, they will be asked to give a blood
sample and to provide written consent for participation and follow-up. Follow-up information
on cause-specific mortality and cancer incidence will be obtained from the office
of national statistics. Data regarding incident morbidity will be received via regular
follow-up hospitalization data and general practice records with confirmation of diagnoses
using standard criteria. Every second year a subset of around 2,000 participants will
be re-surveyed to allow for correction of regression dilution and the entire cohort
will be re-surveyed by postal questionnaire at 5 years to update exposure data and
to ascertain self-reported incident morbidity. The UK Biobank is funded by the Wellcome
Trust, together with the Medical Research Council (MRC) and the Department of Health.
The Coordinating Center is in Manchester and there are several regional centers all
over the country. Pilot studies are being performed now, and it will last several
years until UK Biobank can be used [1]
[4].
To foster collaboration between researchers in the field of population genomics, an
international consortium called Public Population Project in Genomics (P3G) has been
founded. The aim of P3G is the establishment of standards, nomenclatures, communication
tools and sharing of technological know-how. This will allow efficient sharing of
data between projects and with the international human genetics community. Ultimately,
it favours the study of important biomedical research questions that are beyond the
scope of a single effort [7].
It should also be mentioned that there are many scientific arguments for large biobanks
[8]
[9]. However, some scientists have concerns about the upcoming boom. The databases will
only be as good as the individual clinical and exposure information they contain.
Opinions vary whether the standard procedure - a routine examination and a patient’s
health record - is sufficient. A more fundamental critique is, that we already know
that most variation in human disease is due to diet and lifestyle factors, and quantifying
how the risks vary with one’s genetic make-up usually won’t change the solution: encouraging
healthier lifestyles. Thus, the enormous investment in genomic medicine might divert
resources from prevention [10]
[11].
Situation in Germany - POPGEN and KORA
Situation in Germany - POPGEN and KORA
In Germany promising developments took place recently. In the past, the ethical rules
of genetic epidemiological studies were strongly dependent on the local ethics committees
and could be quite restrictive. Meanwhile, the situation is better since common rules
have been agreed upon [12]
[13]. The situation has improved further in 2004, when the German National Ethics Council
has published its Opinion on Biobanks for Research. In this Opinion new and research-friendly
proposals for ethical regulations are made. It is suggested that in the future it
should be possible to perform research without informed consent if the samples and
data are anonymized or pseudonymized. It should be possible that the donor gives generalized
consent for medical research, including unlimited storage. The use of old collections
of biosamples should be possible under specific conditions [14].
Two major German biobanking activities are ongoing. In Northern Germany POPGEN is
being established. The concept of POPGEN is to recruit patients for 8 selected diseases
in the age range of 18 to 80 years in practices and outpatient offices where the patients
are identified from treatment records and diagnoses on the file. The diagnoses will
be verified on the basis of available documentation. It is planed to recruit 15,000
patients affected by one of eight disease phenotypes, and a random sample of 10,000
controls. Regular follow-up of 50 % of the patients is planned [15].
In Southern Germany, KORA [16]
[17] has been used for collaborative genetic epidemiological research, mostly within
the National Genome Research Network (NGFN, [18]) since 2001. From altogether 4261 participants of the last KORA survey S4 (S2000),
2,200 probands have been selected as a control pool for genetic analyses. If more
subjects were needed the whole study group would be at hand. The control population
pool has been used for more than 30 studies on cardiovascular diseases, obesity, type
2 diabetes, allergies, asthma, neurologic disorders, different forms of cancer, and
rare mendelian diseases, as well as for projects dealing with population genetics.
Here, KORA S4 has served to look for population stratification within Germany and
for linkage disequilibrium patterns among European populations [19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32].
Future development - KORA-gen
Future development - KORA-gen
In the framework of MONICA (Monitoring of trends and determinants in cardiovascular
diseases) and KORA (Cooperative Health Research in the Region of Augsburg), four large
population-based cross-sectional studies have been carried out, and a biological specimen
bank was established in order to enable the KORA researchers to perform epidemiologic
research with respect to molecular and genetic factors. The KORA study center conducts
regular follow-up investigations and has collected a wealth of information on sociodemography,
general medical history, environmental factors, smoking, nutrition, alcohol consumption,
and various laboratory parameters. This unique resource will be increased further
by follow-up studies of the cohort.
Now this collection will be opened also for external researchers, under the name KORA-gen
[33]. The objective of KORA-gen is to provide access to information about available population
controls for genetic studies as well as provision of DNA samples, genotypic and phenotypic
data. The KORA-gen infrastructure will be instrumental in questions of study design,
sampling, and matching, of DNA handling and determination of genetic markers, and
of data structures and formats. This will be supported by an internet based information
resource and by providing competent individual counselling and assistance.
A web-portal for genetic control populations will be available. This portal will resemble
modern online libraries that can be searched in great detail. Partners can choose
genetic controls based on age, sex and basic phenotype information. This automatic
pre-filtering will allow a more informed choice of controls that will further be detailed
through individual and person-based counselling.
The biological samples can be genotyped directly at the GSF facilities. Genotyping
also can be performed at other genotyping centers which fulfill predefined quality
and security standards. There will be an amount-dependent fee per sample. The underlying
KORA-gen database will be fed with all relevant information and linked with the KORA
database of the GSF. The static phenotype and biosamples’ databases will be transformed
into a dynamic online system. All genotypes are fed back into a common database for
permanent storage. This gradual accumulation will add significant value to the overall
dataset. In keeping and administrating the central KORA-gen data base, the GSF acts
as a trustee. For all genetic and phenotype data items ownership is defined to those
scientists who provided these data. Access to these data for scientific analyses is
only granted with permission of the data owners. Rules for data ownership and data
access have been formulated and documented as Standard Operating Procedures (SOPs).
The genetic database will adhere to existing standards for data communication with
other partners.
KORA-gen provides data and biosamples of about 18,000 adults from the general population.
It is based on 4 surveys of 4,000 - 5,000 participants each, performed in the city
of Augsburg and the two neighbour counties, with a population of 600,000 inhabitants
[17]. The age range was 25 to 74 years at recruitment and is 30 to 90 years in 2005.
The first three surveys have been part of the WHO MONICA project. The number of participants
is shown in Table 1.
Table 1 Sample size of KORA-gen: n = 18,079 participants of the MONICA/KORA surveys S1 to
S4 in Augsburg. The age range was 25 to 74 years at recruitment and is 30 to 90 years
in 2005
|
survey S1 (1984/85) |
survey S2 (1989/90) |
survey S3 (1994/95) |
survey S4 (1999/2001) |
| men |
2 023 |
2 482 |
2 405 |
2 090 |
| women |
1 999 |
2 458 |
2 451 |
2 171 |
The available data and biosamples are described in Table 2. However, not all parameters are available for all participants.
Table 2 Data and biosamples of the MONICA/KORA surveys which can be used in KORA-gen (some
variables are only available for subgroups)
|
interview
|
medical examinations
|
| - socio-demographic variables |
- systolic and diastolic blood pressure |
| - utilization of the medical health care system |
- anthropometrical measurements (body height and weight, waist and hip circumference) |
| - smoking |
- bioelectrical Impedance Analysis (BIA) |
| - nutrition |
- echocardiography |
| - physical activity |
- electrocardiogram |
| - medication use |
- pulse wave analysis |
| - family history |
|
| - women specific variables |
|
| - self-reported health status |
|
| - psychosocial variables |
|
|
laboratory examinations
|
biosamples
|
| - cholesterol (total, HDL, LDL) |
- serum |
| - uric acid |
- plasma |
| - glucose (partly fasting, partly OGTT) |
- DNA |
| - triglycerides (partly fasting) |
- immortalized lymphocytes |
| - creatinine |
- urine |
| - blood cell counts (WBC, platelets, RBC, MCV, MCH, MCHC platelet volume) |
|
| - HbA1 c |
|
| - haemostasiological and inflammatory parameters |
|
For the surveys S1 to S3 two follow-up interviews with self-administered questionnaires
and mortality follow-up have been performed in 1997/98 and 2002/03. Since 2004, the
KORA study centre conducts regular follow-up investigations of the original survey
population. For details see [16]
[17].
Certain conditions have to be fulfilled when using KORA-gen. The rules set by the
responsible ethics committee and the office for privacy/data protection have to be
followed. Quality standards have to be met with respect to scientifically sound research
questions, study designs have to be based on realistic sample size calculation, quality
of lab tests and genotyping has to be fulfilled according to internationally accepted
standards.
Furthermore the rights and scientific interests of the KORA researchers have to be
taken into account in a fair manner, since they performed the field work and invested
a lot of costs and energy to accumulate the data and biosamples.
The KORA-gen platform will start providing data in spring 2005.
Acknowledgement
Acknowledgement
The investigations have been supported by GSF and grants from BMBF - Federal Ministry
of Education and Research (NGFN: 01GS0423, 01GS0429, 01GR0464, 01GS0485, 01GS0499,
01GS0485, 01GR0411).
The article refers specifically to the following contributions of this special issue
of Das Gesundheitswesen: [17]
[31]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
[41].
