Work

Cohort Profile: The Nijmegen Biomedical Study (NBS)

Year: 2017
Type: article
Study design: Observational Study
Abstract: The Nijmegen Biomedical Study (NBS) is a population-based study that was initiated (by L.K. and A.V.) in 2000 in the municipality of Nijmegen (∼ 150 000 inhabitants) in the eastern part of The Netherlands. The NBS was originally established to obtain a universal reference population to be used for studies of genetic variation, lifestyle and environmental exposures in relation to traits or diseases of interest. However, NBS has also proven useful for studying population traits. The NBS was set up by the Department for Health Evidence, the Department of Laboratory Medicine and the Department of Internal Medicine of the Radboud university medical center (Radboudumc) in Nijmegen in collaboration with the municipality of Nijmegen and the community health service of Nijmegen. At a later stage, the Department of Human Genetics of the Radboudumc joined the NBS project team. Approval to conduct the NBS was obtained from the Radboud university medical center Institutional Review Board. All participants gave written informed consent. The logistic set-up of the NBS was tested in a pilot study performed between November 2001 and February 2002. Via the population registers of the municipality of Nijmegen, the names and addresses of 650 males and females aged 18 years and older were obtained. They each received a questionnaire with questions about, for example, lifestyle and health status. Of the 650, 342 questionnaires were filled out and returned (response rate 53%). Of these 342 persons, 262 (77%) also donated a blood sample. Execution of the pilot study led to further optimization of the study procedure, especially in the blood sampling procedure: the number of community offices for blood donation was decreased and their opening hours were extended. In 2002, the first phase of NBS (NBS-1) was started. On 1 July, a random sample from the register of the municipality of Nijmegen was drawn, stratified by sex and 5-year age groups. Eligibility criteria were age 18 years or older, not living in an institution or rest home and the ability to fill out a questionnaire in Dutch. In total, 22 451 inhabitants of the municipality of Nijmegen were invited to fill out a postal questionnaire (NBS-1 QN) and to donate an 8.5-ml blood sample in a serum separator tube and one (for N ∼ 5000) or two (for N ∼ 1500) 10-ml EDTA blood samples. Of the invited participants, 96% were of Dutch nationality. Table 1 shows the age distribution of the invited participants, separately for males and females. Sampling fractions per sex and 5-year age group are available on request. Age and sex distribution of the sample taken on 1 July 2002 Age and sex distribution of the sample taken on 1 July 2002 The overall response to the questionnaire was 42% (N = 9350), and 69% (N = 6468) of the responders donated blood samples. Table 1 shows an overview of response rates stratified by age and gender. The main reasons for non-participation based on a telephone survey among 65 non-responders were ‘not interested’ (32%), and ‘age (too old)’ (14%). The first phase of NBS has been followed by four additional phases across the period 2002-16 (Figure 1). There are currently no plans to conduct additional rounds of follow-up. Schematic overview of the data collection over time within the five NBS phases. Participants of NBS-1 were invited to participate in subsequent NBS phases if they had given permission to be approached for further research and were eligible for inclusion. FFQ=food frequency questionnaire, N=number, NIMA=non-invasive measurements of atherosclerosis, QN=questionnaire, y=year. NBS-2 was initiated in 2005, with additional health-related questions of interest to the researchers. All participants of NBS-1 who had given permission to be approached for further research were invited to participate. Based on their age on 1 November 2005, participants were invited for different additional methods of data collection, in addition to the basic NBS-2 questionnaire (NBS-2 QN), which was sent to all. Of the 2926 participants who were approached to fill out the NBS-2 QN, 1884 (64%) completed and returned the QN. Of the 2807 participants, 2114 (75%) filled out the NBS-2 QN. Furthermore, 1517 of the 2807 people (54%) participated in the NBS-2-NIMA1 study at the Department of Internal Medicine; this sub-study focused on cardiovascular risk prediction using non-invasive measurements of atherosclerosis (NIMA), namely intima-media thickness (IMT), endothelium function using flow-mediated dilatation (FMD), ankle-brachial index (ABI) and pulse-wave velocity (PWV). All NIMA1 participants (N = 1517) were re-invited in 2012 to fill out an additional questionnaire; 1423 (87%) responded. In addition, 95 of 100 randomly invited NIMA1 participants underwent repeated NIMA at this time. Of these 95 participants, 20 were measured again within 2 weeks to determine repeatability of the measurements. Currently NBS-2-NIMA3 is being executed, with focus on the role of intestinal bacteria in the development of atherosclerosis. Invitations were sent to participants of NBS-2-NIMA2 with a body mass index (BMI) > 27 kg/m2 (N = 561) and to participants with a BMI > 25 kg/m2 and < 27 kg/m2 who indicated that their weight had increased (N = 397); also, friends/family of the participants (N = 145) were invited. Of all invited persons, 302 participated in this part of the study. Of the 2253 participants aged above 70 years who had given permission to be approached for further research and were still alive, 1596 filled out the NBS-2 QN (71%). In 2008-10, NBS-3 was carried out to obtain more detailed information about the nutritional status. A total of 5363 people were invited to fill out a food frequency questionnaire (FFQ). This FFQ was a validated instrument developed by the Division of Human Nutrition at Wageningen University.1,2 Crude questionnaire data were converted to nutrient intake using the Dutch Food Composition Table from 2006 (NEVO 2006); these data are available for 2506 participants (47%). In 2008 NBS-4 was started, to increase compatibility and similarity in available (risk factor) data between the NBS and cancer patient groups that were frequently studied by our research group, as well as to collect trait data and health information for a broader range of studies. A new questionnaire (NBS-4 QN) was sent out to 8109 persons who had given permission to be approached for further research and were still alive, of whom 5613 (69%) responded. The NBS-5-phase was conducted in 2012 in order to collect reference data for a study on risk factors for melanoma and to obtain data on pain, dyslexia and more extensive information on physical activity. Of the 7567 persons who had given permission to be approached for further research and were still alive, 3833 (51%) returned their questionnaire (NBS-5 QN). The NBS-1 QN contained questions on, among other topics, demographics, health status, lifestyle and medical history. The NBS-2 QN covered topics about health and disease, pregnancy, mood and behaviour, daily activities and memory. The NBS-2-NIMA1, 2, and 3 QNs contained questions on general health, medical history (with a focus on cardiovascular traits), use of medication, lifestyle, family history of cardiovascular traits, and quality of life. NBS-3 QN was a food frequency questionnaire. NBS-4 QN was about lifestyle factors and health and disease. In addition, questions about reading problems, mood and behaviour, and life events were included. Finally, NBS-5 QN focused on health and disease, sun exposure, physical activity, pain and reading problems. A more detailed overview of the questionnaire data collected in the different NBS phases is provided in Table 2. Overview of data collected in the different NBS phases. Pdf documents of the questionnaires (in Dutch) and a detailed file with all variables that are available (in both Dutch and English) can be found at [www.nijmegenbiomedischestudie.nl] The first phase of NBS (NBS-1) has been followed by four additional phases (NBS-2, -3, -4 and -5) across the period 2002-2016. NBS-NIMA1, -2 and -3 are sub-studies of NBS-2 focused on cardiovascular risk prediction using non-invasive measurements of atherosclerosis (NIMA). Only for participants older than 70 years. Dutch questionnaire. Validated in a Dutch sample (N = 50) (paper in preparation). Focus on cardiovascular traits. The NBS database has been linked to the database of the Netherlands Cancer Registry and thus also contains official registry data on the occurrence of cancer among the NBS participants; latest linkage covers cancer registry data until 2013. Not for participants older than 70 years Overview of data collected in the different NBS phases. Pdf documents of the questionnaires (in Dutch) and a detailed file with all variables that are available (in both Dutch and English) can be found at [www.nijmegenbiomedischestudie.nl] The first phase of NBS (NBS-1) has been followed by four additional phases (NBS-2, -3, -4 and -5) across the period 2002-2016. NBS-NIMA1, -2 and -3 are sub-studies of NBS-2 focused on cardiovascular risk prediction using non-invasive measurements of atherosclerosis (NIMA). Only for participants older than 70 years. Dutch questionnaire. Validated in a Dutch sample (N = 50) (paper in preparation). Focus on cardiovascular traits. The NBS database has been linked to the database of the Netherlands Cancer Registry and thus also contains official registry data on the occurrence of cancer among the NBS participants; latest linkage covers cancer registry data until 2013. Not for participants older than 70 years In NBS-1, 6468 participants donated a non-fasting 8.5-ml blood sample in a serum separator tube (serum sample), and one (N ∼ 5000) or two (N ∼ 1500) non-fasting 10-ml EDTA blood samples (plasma sample). Blood samples were taken throughout the day; time of blood sampling was recorded. Plasma and serum samples were stored at the Department of Laboratory Medicine at -40 °C until use. Cell pellets were frozen and also stored at -40 °C for future DNA isolation. In NBS-2-NIMA1, fasting blood samples were obtained for all participants: two 8.5-ml serum tubes, two 10-ml EDTA tubes and one 10-ml heparin tube were collected. In addition, a urine sample was collected. In NBS-2-NIMA3, fasting blood samples were obtained again: one 10-ml serum separator tube, two 10-ml EDTA tubes and one 10-ml heparin tube. In addition, urine and faeces samples and swabs from mouth, hand, foot and back for microbiome isolation, as well as adipose tissue biopsies (from abdomen and thigh) were collected. Haematological and biochemical parameters that have been measured in the NBS blood samples are presented in Table 3. Overview of haematological and biochemical parameters that have been measured in the NBS blood samples The first phase of NBS (NBS-1) has been followed by four additional phases (NBS-2, -3, -4 and -5) across the period 2002-16. NBS-NIMA1, -2 and -3 are sub-studies of NBS-2, focused on cardiovascular risk prediction using non-invasive measurements of atherosclerosis (NIMA). Samples were non-fasting and taken during the day. Time of blood sampling was recorded. Fasting samples taken in the morning. Transferrin saturation is calculated by dividing serum iron by total iron-binding capacity. Hepcidin and macrophage migration inhibitory factor have been measured in 2998 samples. Overview of haematological and biochemical parameters that have been measured in the NBS blood samples The first phase of NBS (NBS-1) has been followed by four additional phases (NBS-2, -3, -4 and -5) across the period 2002-16. NBS-NIMA1, -2 and -3 are sub-studies of NBS-2, focused on cardiovascular risk prediction using non-invasive measurements of atherosclerosis (NIMA). Samples were non-fasting and taken during the day. Time of blood sampling was recorded. Fasting samples taken in the morning. Transferrin saturation is calculated by dividing serum iron by total iron-binding capacity. Hepcidin and macrophage migration inhibitory factor have been measured in 2998 samples. Initially, genome-wide genotyping was carried out for 1980 samples using the Illumina HumanHapCNV370-Duo BeadChip platform, of which 1819 survived quality control (QC). Currently, we have genome-wide genotype data measured with the Illumina HumanOmniExpress-12 and -24 BeadChip available for 5363 samples; 5292 of these passed a call rate threshold of 95% and were imputed using the 1000 Genomes phase1 v314 together with Genome of The Netherlands (GoNL) release 515 data as reference. Pre-imputation QC on the marker level consisted of a minor allele frequency (MAF) > 0.01, Hardy-Weinberg equilibrium (HWE) P-value > 10-4 and a single nucleotide polymorphism (SNP) yield > 95%, resulting in 609 046 SNPs to be used in the imputation process. Imputation was performed using the Impute2 pipeline developed by the GoNL team; see [http://www.bbmriwiki.nl/wiki/Impute2Pipeline].16 This resulted in 20 011 335 SNPs. Post-imputation quality control consisted of exclusion of population outliers using principal component analysis, exclusion of sex discrepancies based on a comparison of genotype data and clinical data, and a relatedness check, which resulted finally in 4745 samples available for genome-wide analyses. In addition, 1931 NBS samples have been chipped with the Illumina HumanExome BeadChip, which covers putative functional exonic variants and is focused on the measurement of rare (MAF < 0.5%) and low-frequency (0.5%< MAF < 5%) variants [http://genome.sph.umich.edu/wiki/Exome_Chip_Design]. A total of 242 901 variants were measured with this chip, and data were called both in GenomeStudio and with zCall, a caller specifically designed for calling rare variants.17 After quality control, that is sample call rate > 95%, removal of sex discrepancies, a relatedness check, a heterozygosity check, exclusion of population outliers, marker call rate > 98%, exclusion of markers that should be treated with caution–see Exome Chip Design Wiki: [http://genome.sph.umich.edu/wiki/Exome_Chip_Design]–and HWE > 10-6, a set of 1825 samples and 242 703 markers are available for association analysis. The NBS database has been linked to the database of the Netherlands Cancer Registry in order to obtain official registry data on the occurrence of cancer among the NBS participants. Linkage was executed based on identifying information such as name, date of birth and date of death. Linkage has been repeated several times in order to update information on the occurrence of cancer. The latest linkage was performed in 2014, resulting in complete information on the occurrence of cancer in the NBS participants until 2012 and partly for 2013. Periodically (in the beginning monthly, nowadays every 6 months), we receive updates from the municipality of Nijmegen about changes in vital status or address of the NBS participants who still live in the original catchment area. In this way, we have an up-to-date database of the latest contact information and vital status of the NBS participants. This linkage does not provide information on causes of death, neither can we obtain these data due to a lack of informed consent. The first paper based on NBS data was published in 2006, which was about thyroid function and presence of anti-thyroperoxidase antibodies in the NBS.18 Up to 22 June 2016, there have been 171 peer-reviewed articles published that used data from the NBS. An up-to-date list can be found on our website [www.nijmegenbiomedischestudie.nl]. The papers cover a wide range of topics, as there is a large variety of demographic, clinical, biochemical and genetic variables available. The project team of the NBS itself is focused on research into bladder cancer, atherosclerosis and cardiovascular risk factors, disorders of iron metabolism and psychiatric disorders in the adult population (like attention deficit/hyperactivity disorder (ADHD) and autism spectrum disorders). The NBS has contributed genetic data to many international consortia that studied genome-wide DNA variation in relation to binary disease phenotypes, using the NBS participants as a control group, and quantitative traits. This has led to the identification of novel associations between DNA variants and a whole range of traits and diseases, for example urinary bladder cancer,19–22 schizophrenia23 and educational attainment.24 Furthermore, smaller-sized genetic studies that zoomed in on specific candidate genes have been performed using NBS data.25–27 Also, genetic variation has been used to draw conclusions about causality of risk factor-outcome associations. An example is a Mendelian randomization study that indicated that iron traits might play a role in atherosclerosis.28 Phenotypic data from NBS have also been used in many epidemiological studies. Reference values have been constructed for thyroid function,18 glomerular filtration rate29,30 and hepcidin.31 In addition, data from the NBS-2-NIMA sub-study have been exploited to study risk factors and risk prediction models for atherosclerosis and cardiovascular traits.32–37 This revealed, for example, that waist circumference is independently associated with subclinical atherosclerosis.36 Finally, NBS has been used as a control group for studies into risk factors of bladder cancer and prostate cancer.38–40 One of these studies showed that there is no association between personal hair dye use and bladder cancer risk, also when taking various types of hair dye, intensity of exposure to hair dyes or dye colour into account.38 The NBS was designed to obtain a universal reference population and it has been extremely valuable as such. A main strength of NBS is that the participants are well with data on a wide variety of environmental and NBS thus a of information on a range of research In addition, data collection in phases has collection of information on additional topics that were not covered in the NBS database has been linked to of the Netherlands Cancer the of data available for The NBS is also a large study questionnaire data on 000 participants and data for more than participants. Finally, contact information and vital status of the NBS participants living in the original catchment have been up to date using information on from the register of the municipality of Nijmegen which is sent to on a NBS participants are thus for future research as a population for new studies. example, NBS is currently in the Netherlands study NBS participants were invited and have All participants of the NBS were inhabitants of the municipality of Nijmegen at the time of inclusion. the NBS the population of the eastern part of The Netherlands. well the NBS the population and the Dutch population has not been is that participants in the NBS are based on with the level of the total Dutch population and of that are being studied by the NBS project team. In addition, it is to that the age distribution of the NBS is not of the age distribution of the population of the municipality Nijmegen, as the NBS sample was from the register of the population of the Nijmegen stratified by 5-year age groups. In order to draw conclusions about of diseases, the sampling for each age group should thus be taken into Sampling fractions are available on request. the NBS is a and not a The NBS is older over and of participants in NBS due to of the In addition, the of participants who contributed to later phases of that is NBS-2 to a group of which should be taken into when about the of study to the general population using data from later phases of NBS. Finally, blood samples collected in NBS-1 were taken non-fasting and not at a time during the day. This is for of the biochemical such as the blood and the iron can be taken into by using time of blood sampling in analysis. The data of the NBS are available to the international The website [www.nijmegenbiomedischestudie.nl] contains information on the collected data, documents of the questionnaires (in Dutch) and a detailed file with all variables that are available (in both Dutch and can for use of the data data, parameters and cancer registry linkage for by a research to which should information on the of the the research questions to be an and the data be by the NBS project and a be the which the data be The data be provided of be used to the research in the research In the NBS collection was to the Radboud an within the Radboud university medical center for the collection, and of and the to clinical The are also available to the community on a can be to the Radboud see for the The NBS is a population-based study that was established to obtain a universal reference population that can be used in a variety of studies in order to study genetic variation, lifestyle and environmental exposures in relation to a variety of traits or The first phase of NBS was in 22 451 inhabitants of the municipality of Nijmegen were of whom filled out the questionnaire and 6468 donated a blood sample. The age from 18 to years. The phase of NBS has been followed by four additional phases from to 2016, additional questionnaires and clinical and biochemical measurements. information and vital status of the NBS participants living in the original catchment have been up to date until using information from the register of the municipality of more than of the participants are still for future studies. data a wide range of environmental and biochemical genome-wide genetic information and linkage to the Netherlands Cancer Also, and DNA can be are available to the international a to see for more The set-up of NBS was by the four Radboudumc and a from the Radboudumc Board. The municipality Nijmegen and the community health service Nijmegen also contributed of to all participants of the NBS. In addition, we and for their to the NBS.
Language: English
Subfield: Genetics
Open Access status: hybrid
Estimated APC paid: $4,514
OpenAlex ID: W2576685034
FWCI: 2.845
Cites: 42
Cited by: 45
Related to: 10