Data Resource Profile: The ALSPAC birth cohort as a platform to study the relationship of environment and health and social factors

This resource profile describes the information about the physical and social environment collected within the Avon Longitudinal Study of Parents and Children (ALSPAC) birth cohort. This includes spatial and temporal information gathered on three generations about: area-level built and social characteristics (e.g. density and location of fast-food outlets, crime rates within a neighbourhood); exposure measurements (e.g. air pollution concentrations, temperature records); participant-reported data directly related to the spaces and places they inhabit (e.g. neighbourhood safety, presence of damp within a home); information directly measured from participants (e.g. blood lead and total mercury concentrations, physical activity); the location information needed to link these diverse data. We describe the platform’s previous uses, strengths and weaknesses and access arrangements, emphasizing confidentiality safeguard controls. This profile highlights a particular class of ALSPAC data (with distinct access arrangements) to promote the potential for incorporating physical environment and other spatially-dependent data into research investigations. ALSPAC is a multi-generational prospective birth cohort study1,2 that has compiled an exceptionally detailed longitudinal resource of directly measured and linked phenotype and ‘omic’ data.3 ALSPAC’s eligible sample is defined as all pregnant women living in and around the city of Bristol (south-west UK) and due to deliver between April 1991 and December 1992. Women carrying a total of 20 248 pregnancies were deemed eligible. Of these, ALSPAC has recruited ‘G0’ (generation zero) mothers of 15 247 pregnancies, which resulted in 15 458 ‘G1’ (index generation) fetuses. Of this total sample of 15 458 fetuses, 14 775 were live births and 14 701 were alive at 1 year of age. By April 2018, the G1 index participants had reached young adulthood, with many having children of their own. Recruitment of the third-generation ‘G2’ (children of the G1 index sample) began in 2012 and included recruitment at any age from in utero onwards.4 By January 2019, over 907 G2 children from over 604 families have been recruited into ALSPAC. The ALSPAC catchment was centred around the city of Bristol, 106 miles west of London. It comprised three health administration districts within the South-West Regional Health Authority that later became the ‘Bristol & District Health Authority’ (Figure 1). This area largely overlapped with the County of Avon, which was restructured in 1996 into the City of Bristol and the counties of Bath & North East Somerset, North Somerset and South Gloucestershire. Local employment is largely tertiary sector (i.e. commercial and government services), although Bristol is noted for high-tech aerospace manufacturing and agriculture and food/drink production. The area has a temperate climate and its geology is predominantly sedimentary (carboniferous limestone). Local natural resources include coal, iron, lead and zinc, which have been mined locally for up to 2 000 years.5,Table 1 describes differences between the City of Bristol, the wider metropolitan area and the whole of England and Wales in terms of population growth, density, age, ethnicity and economic activity. More detailed demographic and environmental assessments are compiled in UK government reports6,7 and census-based reports.8–10 The ALSPAC Eligible Study Area within the UK: illustrating the NHS District Health Authorities (DHAs) used to define: the ALSPAC catchment area; the historical county of Avon; and the four authorities formed following the breakup of Avon. Contains Ordnance Survey, Office of National Statistics and National Records Scotland data © Crown Copyright/database right 2014. Comparison of a selection of population characteristics in the City of Bristol, the wider metropolitan area and the whole of England Source: Office for National Statistics (ONS). ONS Crown Copyright Reserved. Data in the ALSPAC resource can be linked with physical and social environment records using geocoded databases recording participant life-course location. Central to ALSPAC acting as a platform for geospatial and temporal research is the database of participant addresses and other key location information (e.g. school addresses). ALSPAC has maintained an administrative database of participant address details since recruitment, and invests considerable resource in maximizing the completeness of this record (see Supplementary materials, available as Supplementary data at IJE online). Participants’ address records have been cleaned and geocoded at the property central coordinate (centroid) and postcode centroid level to 1-m Easting and Northing cartesian coordinates. In the UK, postcodes group properties into small clusters containing an average of 15 properties (range 1–100). We used coordinate data in Ordnance Survey’s AddressBase Plus product to geocode properties, and the Office of National Statistics Postcode Directory (ONSPD) to geocode postcodes. G1 school attendance records from the English National Pupil Database (NPD) attainment and census databases were combined to create a record of school attendance from age 4/5 to 18 (∼85% of G1 index participants are linked to their NPD record). School postcode centroids have been geocoded to 10-m coordinates (see Supplementary materials, available as Supplementary data at IJE online). This geocoded database will be iteratively updated over time and could be extended to new participant-related locations (e.g. registered general practice, workplace) or, in theory, dynamic location models based on personal sensor data. The ALSPAC database contains participant self-reported, teacher-reported and study fieldworker-assessed variable data and biological measures related to the home, school and natural and physical environment (including participant neighbourhood). These data have been collected at individual and household levels and are summarized in Table 2. G0 mother-reported data about the home they and their G1 child live in (e.g. degree of damp and mould, new furnishings and carpets) have been collected at six time points: antenatally, and at G1 age 8 months, 21 months, 33 months, 61 months and 85 months. There is also self-reported information on modes of transport and occupational history of the G0 mother and her G0 partner. ALSPAC and other fieldworkers have collected air pollution data from subgroups of the G0/G1 homes using sensors (e.g. Palmes NO2 sensor) and through collection of biological samples (e.g. venous blood carboxyhaemoglobin and methaemoglobin readings). Biological measures of lead, cadmium and total mercury and other elements were measured in G0 maternal blood early in pregnancy, in G1 umbilical cord tissue samples (samples collected by midwives) and in G1 children’s blood (for lead only) at age 30 months (collected in the Children in Focus assessment clinic). ALSPAC is testing ‘internet of things’ approaches11 and the use of personal air pollution sensors and location tracking sensors (unpublished), and is collecting information using head cameras to study interaction between G1 parents and their G2 offspring at home, in the garden and at other locations.4 Summary of ALSPAC participant reported and study collected measures Formaldehyde, toluene and other volatile organics Nitrogen dioxide Fungi, house dust mite and bacteria Temperature and humidity Antenatal clinic ICP-DRC-MSb Maternity hospital ICP-OESd/ICP-MSe CO indoor background (Draeger diffusion tube) CO exhaled breath (Bedfort EC50 ToxCO breath CO monitors) Variables including: Type of housing, including storey Degree of damp and mould in each room Frequency with which windows were opened in summer/winter Type of heating and cooking used Wallpapering, painting, new furniture or carpets and in which rooms Household/occupational/hobby chemical use Noise Variables including: Traffic density on the road Modes of transport Time spent outdoors Mothers and fathers/partners occupational history (coded to SOC90) neighbourhood quality Variables including: Distance to road Noise Building and facility quality m, months. The Building Research Establishment (BRE) study was of 174 homes (quasi-random selected) and each assessed over a 12-month period. ICP-DRC-MS, inductively coupled plasma dynamic reaction cell mass spectrometry. In addition: selenium (Se). ICP-OES, inductively coupled plasma optical emission spectrometry. Elements (except Pb) were assayed by ICP-OES (n = 2005), except for Se and Hg, which were measured by atomic fluorescence techniques (hydride generation and cold vapour, respectively); the final 911 samples were assayed for these elements plus Pb by ICP-MS. In addition: Se, Mg, Ca, Cr, Mn, Fe, Co, Ni, Cu, Zn, Sr, Mo, Sb, K. AAS, atomic absorption spectroscopy. The ALSPAC geocoding can be used to link any information recorded against a geolocated point or area into the ALSPAC databank. Whereas the possible linkages are numerous, these fall under a range of broad data domains (e.g. geological, housing stock characteristics) that can inform assessment of a range of exposure types (e.g. ambient residential exposures such as potential domestic radon gas exposure from underlying geology) (see Table 3). Domains of geolocated information and types of exposure that are, or could potentially be, linked to and evaluated with ALSPAC data Meteorological, climate and associated emissions (e.g. ultraviolet radiation) Outdoor ambient air quality (e.g. industrial air pollutants, pollen count) Water quality (e.g. drinking water additives and quality) Green space, blue space and land use (including plant species) Geological (e.g. radiation) and topographical (e.g. altitude, aspect and hydrology) Noise, vibration, radiation and electromagnetic fields Ambient residential exposure (e.g. air pollution, noise levels) Ambient occupational exposure (e.g. noise levels) Indoor residential exposure (e.g. indoor air pollution) Modelled commute exposure (e.g. air pollution, pollen count) Other residential exposures (e.g. water quality, radiation) Accessibility to services and amenities (e.g. green space) Extent and density of built environment Active transport connectivity These data can be collected in situ (local point measurement, such as air pollution sensor data). Alternatively, the data can be modeled using: survey data (e.g. maps of radon potential); satellite remote sensing data (e.g. vegetation measured from land cover type); spatio-temporal data (e.g. time-resolved air pollution exposure maps); or environmental exposure proxy data (e.g. distance to nearest main road; traffic intensity (count) multipled by distance to the nearest main road). Consequently, records can be generated locally, nationally or internationally and by a diverse range of organizations. Table 4 provides illustrative examples of these data sources and includes a more detailed exemplar summarizing the potential linked data inputs that could be used to model NO2 exposure in the ALSPAC catchment area from road and local sources using a methodology established previously to model particulate matter exposure.12 Illustrative examples of physical environment data that could be linked to ALSPAC, including a summary of the potential sources to inform NO2 modelling National ‘static’ maps and inventories DEFRA annual average background air pollution maps national atmospheric emissions inventory (NAEI) Time-varying, spatially-gridded validated governmental / agency data Met Office meteorological data ECMWF CAMS modelled atmospheric data Nationally distributed time-resolved point measurement data DEFRA AURN measured air quality data CEH COSMOS-UK soil moisture measurement network data Local government Bristol environmental survey County road traffic Research data measurement, modelling and in data by and satellite data optical data the natural and physical environment air pollution model models exposures from sources use models of the built and natural and physical environment Building Data National 30 data from model ECMWF CAMS DEFRA air pollution in situ point measurements for National on annual average pollution sources and emissions County road traffic data City historical measured diffusion data on NO2 exposure over and ALSPAC data on Bristol City data can be traffic data can be research data can be postcodes the UK) have been UK using the National Statistics Postcode which provides health and at as as neighbourhood and index and their and of These have been to all data collection G1 and G2 materials, available as Supplementary data at IJE online). of these can be used for and the can be used to link national and locally collected data such as the Bristol of to neighbourhood characteristics or to or ALSPAC can and into and exposures or (e.g. distance to nearest ALSPAC has an of and at and with to and health records and social administrative The can be the study records can be data a variable and such as ALSPAC to new including environmental data. data collection can potentially use and and (e.g. study assessment biological and could include to records or the ALSPAC new data collection The to ALSPAC through has established a data model for and data into combined data (Figure any data the model a distinct that data using that to and maximizing potential for use through as many data as possible on as a of the ALSPAC sample as a that use of to link data to and data that and have and data model illustrating the of radon exposure their to cohort participants using the the of ALSPAC has the for as (i.e. the by which exposure data are and into data which are with the wider ALSPAC This for all longitudinal population by the Area Health Statistics and ALSPAC to link geolocated data to participants and exposures at key at an individual the of participant (e.g. and address exposures and exposure and exposure over ALSPAC has a for cleaned address and and can data quality to research as data data and cleaned cleaned addresses). It is that ALSPAC’s data is and to participants and is with We have participant to participant on the use of spatial data in ALSPAC research and are on this ALSPAC data the to participant on personal location this research is as and within the of the and participants had or to be to this of in around 30 participants In data an to were to the data research that location (e.g. 1 locations (e.g. home or school and location (e.g. participants summarized for this with this by the of the was that this of research is the potential to health was Research using personal location data was as from other within of the participants that the data that have been collected be the of the could be by that are in for other types of ALSPAC for for data for and of data. There was around the of to ALSPAC the participants personal of and the be by wider small of participants about of and location data. In the group were with the of location data and this was as as personal as the other location data under a participants about the potential for cell were to location was that locations are as more participants that they were more for their school address to be their home to the of other at the school the of small cell There was that the of locations the of and that locations as is as they be for participants and their The in to location data was that could be linked through more Of this is to ALSPAC, also to many other longitudinal participants that be access to these data. this a for a small of the group was with to collecting and (e.g. data. participants this It was as this new data participants could to in participant that new data collection be by an and that their more in the access to these data. were by the group as the use of these data in was a general that many participants these in their to commercial collection of location data. was that participants this of data collection more of with this of data In general participants were by (e.g. that they a natural had about the location (e.g. that they were on a particular road they participants had particular to these data to their the in these location data and as within and was for participants to or the with key to the the of using a in the main was and a of to be a general for in research that the personal location data to be at ALSPAC, was also of the this in participants to that this of research is in a of research of was with an and of as in with previous in ALSPAC (e.g. for by the to which personal location data such as addresses are as data as a for to be was that the types of research by use of personal location data be and within for the of participants to that use of personal location data was the that could be in and that the the between that the that are in could in on a The of and personal location data was more data. the that participants are at to this the was a of in ALSPAC by its Participants’ have been to the data access for location data and The access includes around the ALSPAC which social (e.g. data access information and (e.g. The is for ALSPAC data to with in a that the participant postcode data are the broad of postcodes are or information from these (e.g. household of of at the time of data that are this information is for such as for spatial research In these will with the to their the or using a of the as The will be with a containing postcode and any other data. the of the participant postcodes will be by including postcodes and data. The will use this to to The will of the to the Study all of the 1 data. The ALSPAC Data the to the ALSPAC the and this to the The will be for and be to a level to this will be and in to for spatial for be in is be at address or postcode data will be under any a range of administrative are available as in the data will be on its and of data are a of a of a that includes ALSPAC spatial a will be to include detailed information on the and methodology the to the and to the spatial is for the research in the of the as to data are Data with the many to due to population with the possible can be to The of address data has a database with temporal due to historical database and individual level differences in address will be a small of that have address data at time These research that ALSPAC as is a between address and for the ALSPAC study was from the ALSPAC and and Health Research Authority research of the ALSPAC have been on geocoded data or the use of geospatial and many have used information as (e.g. for social using or have used to modelling (e.g. include (for examples Supplementary materials, available as Supplementary data at IJE between domestic exposures and maternal and child health household chemical product use and child and NO2 from household sources and health electromagnetic radiation exposure to pregnant mothers that exposures from were by the of the home be measurement within the of of lead, mercury and cadmium exposure with of child cord and of child blood lead with school has been used to of a of exposures to a key child the of particulate matter air pollution exposure on that exposure in early of the on the by ALSPAC of in blood lead and selenium have have the potential to the environmental for to was linked to and an between a in the to of the maternal between health and These an between and and that maternal were associated with birth the of residential location and residential on health and social residential and the of underlying to previously between residential and residential and and the of on residential between places (e.g. the from home to with fast-food and the of of on to environmental exposure modelling techniques within longitudinal research including modelling particulate matter exposures and neighbourhood measures (e.g. used to inform in and ALSPAC phenotype data that have been to inform local health and ALSPAC the in which neighbourhood are the of the quality of survey and and information from study and testing the of collecting exposure data within an The of this resource is ALSPAC’s to link data to the ALSPAC databank. geocoding the from assessment of in utero to history has been by school location and could be extended to other ALSPAC research through of information and through linkages to new data. The needed to participant confidentiality could be a they on data research to a of data participant ALSPAC’s is as participant provides to (e.g. local and is to such as modelling using and area ALSPAC in be to to national quality on the quality and completeness of participant location which is participants are to that to is is that participants with the dynamic history (e.g. in or to have quality location data. these quality is directly data is have their and the weaknesses are to through tracking and (i.e. collection of location the collection of address information through record and the potential to use to address at The ALSPAC is as a resource for the research data are the of data use is and is the to of or the ALSPAC data access and for access geolocated data (e.g. for are available at each data time of data are available the UK Data linkages of information ALSPAC data linkages are assessed for with ALSPAC’s and data use Data are to generated or data with for in ALSPAC. by information and and participant confidentiality are for to a ALSPAC the right to to to address to participant or the study into The ALSPAC birth cohort study has compiled a detailed resource of physical and social environment measures needed to between environmental exposures and health and social ALSPAC is a cohort. eligible all pregnant women living in and around Bristol, UK, and due to deliver April their (including the biological the index children (n = 15 247 the index offspring they have an the index data have been collected using: measures by participants and associated with (e.g. participant household assessments assayed biological and linked based on geocoded participant address Data indoor and air pollution noise neighbourhood quality linked and census and data can be linked from records and modelled exposures (e.g. meteorological data). data are available to the ALSPAC The UK Research and and the of Bristol for ALSPAC of ALSPAC is available on This was by the UK Research as of the data In ALSPAC The the wider of the ALSPAC geospatial resource through the to ALSPAC through ALSPAC G2 data collection and are by (see and also the Research under the and National of and and and in or are to a that from the UK Research and of is a is by the of Bristol Research and within the is by a of for environmental data. This is the of the will as for the of this We are to all the families in ALSPAC, the for their in and the whole ALSPAC which includes and research and We to the of the ALSPAC for their into ALSPAC’s for data. of