All time references are in CEST
New approaches to biomeasure collection in social surveys |
|
| Session Organisers | Ms Anne Conolly (National Centre for Social Research ) Dr Erica Wong (CLS, UCL Institute of Education) Dr Colter Mitchell (University of Michigan) Dr Jonathan Burton (ISER, University of Essex) Mrs Christine Woods (ISER, University of Essex) Dr Shaun Scholes (UCL) |
| Time | Tuesday 18 July, 09:00 - 10:30 |
| Room | |
The inclusion of biomeasures is an important feature of many social research surveys, providing the objective measurement of a wide range of health indicators. Recent shifts in survey research methodology, for example remote and mixed-mode approaches, pose challenges for the successful implementation of biomeasure data collection, including development of new approaches.
Biomeasures include a range of anthropometric (e.g. height, weight, waist), functional (e.g. grip strength, balance), and sensory (e.g. hearing) measurements, as well as biological samples (e.g. blood, saliva, urine), other physiological health measurements (e.g. blood pressure, lung function), and device-based measurement of physical activity.
Historically, biomeasure data has primarily been collected in participants’ homes or at clinics. Data collection is commonly administered by those with medical training and expertise (such as nurses), by trained field interviewers or self-administered by participants. The growing trend towards remote survey methods requires the development of new approaches in biomeasure data collection. These include biological samples by mail, app-based data collection, device-based measurement and the application of correction factors to self-reported anthropometric data. These alternative approaches often present survey researchers with a range of challenges that present risks to data quality.
This session invites survey practitioners to share their experiences of developing and implementing new approaches to biomeasure data collection. We welcome submissions relating to:
• Innovative approaches to the collection of biomeasures
• Comparisons of different biomeasure collection approaches
• Respondent-led collection of biomeasures
• Maximising response to and/or representativeness of biomeasures
• Maximising data quality and validity of biomeasures
• Ethical challenges in the remote collection of biomeasures (e.g. relating to consent, feedback of results, etc.)
Papers need not be restricted to these specific examples.
Keywords: biomeasures, health data, biosocial, biomarkers
Mrs Beverley Bates (National Centre for Social Research) - Presenting Author
Mr Kerry Jones (Epidemiology Unit, University of Cambridge)
The National Diet and Nutrition Survey (NDNS) is the only nationally representative survey on diet and nutritional status of the UK population. Whilst NDNS started in 2008 as a traditional in-person survey it moved in November 2024 to a web-first methodology, which vastly increases the amount of data and the speed at which it is collected. A key NDNS measurement is urinary iodine concentration measured in spot urines of those aged 4 years and older. The aim is to collect and analyse samples from 800 participants across the UK per year.
In the presentation we will give an overview of the practicalities of collecting remote samples in multi-participant households. We will explore whether the representativeness of those providing a spot urine sample and the quality/usability of the collected samples are comparable to the data previously collected via in-person visits. We will also consider whether a remote protocol can be extended to other measures collected in NDNS and further, how much complexity is realistic to expect a web- first design to handle.
Dr Lauren Gaydosh (UNC Chapel Hill) - Presenting Author
The aging population in the United States is increasingly diverse, particularly with respect to gender identity and sexual orientation. Despite their increased risk for poor aging outcomes, we have very little data on biological aging among this population. We detail the protocol for the self-collection of biospecimens among a pilot sample of respondents from the LGBTQ+ Social Networks, Aging, and Policy Study (QSNAPS). Participants provided buccal DNA and capillary blood which were used to measure DNA methylation and blood protein measures of aging, respectively. We use these data to construct measures of biological aging and interrogate patterns by gender identity and sexual orientation. These data provide a first look at the biological aging of sexual and gender minority older adults.
Dr Jonathan Burton (ISER, University of Essex) - Presenting Author
Professor Michaela Benzeval (ISER, University of Essex)
Mrs Christine Wood (ISER, University of Essex)
Understanding Society: The UK Households Longitudinal Study is a large-scale annual study that started in 2009. After Waves 2 and 3, a sub-sample of the study participants received a nurse visit to collect a range of biomeasures. In 2024, we launched Wave 16 of the study, which was also a biomeasure wave. Whilst starting as face-to-face, the study has shifted to a mixed-mode survey that uses web, face-to-face, and telephone interviewing. The move towards a mixed-mode survey, with more participants completing their annual interview online, required us to develop new methods of collecting biomeasures for Wave 16.
This presentation describes development and testing process that influenced the design of Wave 16. It covers a range of experiments using the Understanding Society Innovation Panel and the pilot for Wave 16. It includes the comparison of eliciting biomeasures from those who complete online with those who are interviewed in-person by a field interviewer or a nurse. We have tested the use of two apps to measure (i) body volume, and (ii) spatial awareness as a potential early sign of dementia. We have also tested the collection of biomarkers using dried blood spots, capillary blood, and a stool sample to collect microbiome. We investigated both the uptake of different protocols and measurement errors.
Wave 16 of the study launched in January 2024, and will continue until May 2026. As well as presenting information about the development and protocol of Wave 16, we will also present some early results on the take-up and return of biomarker samples.
Dr Erica Wong (Centre for Longitudinal Studies, University College London) - Presenting Author
Professor Pasco Fearon (University College London and University of Cambridge)
Professor Lisa Calderwood (Centre for Longitudinal Studies, University College London)
Professor Alissa Goodman (Centre for Longitudinal Studies, University College London)
Ms Karen Dennison (Centre for Longitudinal Studies, University College London)
Dr Alyce Raybould (Centre for Longitudinal Studies, University College London)
Dr Konstantinos Tsigaridis (Centre for Longitudinal Studies, University College London)
In longitudinal birth cohort studies, collecting DNA samples enables researchers to understand the interplay between biology and social environment and doing so at the baseline wave of a study provides an opportunity to genotype children and parents before subsequent attrition takes place. However, collecting saliva samples for DNA extraction at the first wave has not yet been tested in the UK, and it is unclear what the impact on overall recruitment to a longitudinal study would be. In order to test the impact of saliva collection on recruitment at an initial wave, a randomised subsample of families was asked to provide saliva for DNA extraction in the ELC-FS.
The Early Life Cohort Feasibility Study (ELC-FS) collected information about several thousand babies aged between 8-12 months old, and their families in 2023-2024 in order to test the feasibility of conducting a new UK-wide birth cohort study. It is led by the Centre for Longitudinal Studies at University College London with interviewing carried out by Ipsos. Interviews were carried out using web, phone and video-interviewing as well as face-to-face, and mothers and fathers (including those living in their own households) were recruited directly and separately.
In ELC-FS, saliva consent and collection could be administered in-person or remotely, and saliva was collected (if consented) for each biological parent as well as the cohort child. We discuss the qualitative testing conducted ahead of fieldwork to test materials and acceptability of this, and we present results on the survey response rates of the saliva subsample compared to the non-saliva sample, as well as the consent and return rates for parents and babies, and variation in those rates across key sociodemographic subgroups.
Mr Tim Weihrauch (Robert Koch Institute, Department of Epidemiology and Health Monitoring, Berlin, Germany / Charité Universitätsmedizin Berlin, corporate member of Freie Universität Berlin and Humboldt-Universität zu Berlin, Germany) - Presenting Author
Mr Ilter Öztürk (Robert Koch Institute, Department of Epidemiology and Health Monitoring, Berlin, Germany)
Mr Lemcke Johannes (Robert Koch Institute, Department of Epidemiology and Health Monitoring, Berlin, Germany)
Dr Susanne Krug (Robert Koch Institute, Department of Epidemiology and Health Monitoring, Berlin, Germany)
Dr Kristin Manz (Robert Koch Institute, Department of Epidemiology and Health Monitoring, Berlin, Germany)
Mrs Marie Rissom (Robert Koch Institute, Department of Epidemiology and Health Monitoring, Berlin, Germany)
Mrs Charlotte Kühnelt (Robert Koch Institute, Department of Epidemiology and Health Monitoring, Berlin, Germany)
Dr Ronny Kuhnert (Robert Koch Institute, Department of Epidemiology and Health Monitoring, Berlin, Germany)
Mr Daniel Grams (Robert Koch Institute, Department of Epidemiology and Health Monitoring, Berlin, Germany)
Dr Julika Loss (Robert Koch Institute, Department of Epidemiology and Health Monitoring, Berlin, Germany)
Objective: Sensor-based measurement of physical activity (PA) and sleep behavior, e.g. using wearables, can complement standardized surveys in national health monitoring. Wearables provide more accurate data than subjective reports, but data from user-owned devices may not be representative, as adults who own wearables are typically more physically active than the general population. Sending wearables to a representative sample can overcome this selection bias. We conducted a pilot study testing the feasibility of (a) mailing wearables to participants of a national panel (“Health in Germany”) and instructing them to collect data for two weeks, (b) feeding the collected data into the panel database. Thus, we intended to assess whether the study design is suitable for broader application in national surveillance by the RKI.
Methods: Panelists were invited to participate by e-mail. Study materials, including a commercial wearable, instructions, account details, and further information, were sent by post in 12/2024. Participants are required to download two apps for data transfer. A follow-up survey is conducted via one app. Participants are asked to return the wearable after 2 weeks. We evaluate response and dropout rates, participants’ feedback and engagement, and wearable data.
Results: Of the 793 panelists invited, 233 consented within two days. Devices were mailed to 147 panelists. Early observations highlight challenges in coordinating service providers and technical issues (password management, QR code usability). As the study is ongoing, results will be presented at the conference.
Conclusions: The use of wearables represents an innovative approach to collecting biomeasures in surveys, enhancing accuracy and reducing bias compared to self-reported data. High data protection requirements complicated data collection procedures for the participants and compromised feasibility and acceptability. By identifying challenges and solutions, this study advances methods for integrating sensor-based biomeasure collection into large-scale surveys.
Dr Jessica Faul (University of Michigan) - Presenting Author
The impact of returning medical or biological results to participants on their consent and participation has rarely been tested in large, representative studies. Collecting physical and biological data is costly and burdensome, so understanding whether returning results can boost consent is crucial for data collectors.
Offering to return results seems a straightforward strategy to incentivize participation. Feedback might motivate more people to consent to the collection and may help keep them engaged in the study for longer, thus improving response rates to follow-up surveys. However, this approach is costly and raises ethical questions about the information provided. It can also potentially alter observational studies into interventional ones.
This project reports on an experimental study embedded within the 2024 wave of the Health and Retirement Study, testing the effects of returning personalized reports from linked administrative databases on pharmacy history and clinical lab test results. ExamOne, a subsidiary of Quest Diagnostics, offers linkages to nationwide databases of pharmacy and clinical lab results. ScriptCheck provides a seven-year pharmacy history with a 90%+ match rate, while LabPiQture offers up to seven years of clinical lab results with a ~70% match rate. Both require prior HIPAA-compliant consent from participants.
Participants are asked to consent to these linkages in the current HRS wave, with an anticipated 80% consent rate. The default treatment does not offer result returns. We have selected case and control primary sampling units matched on demographics to test the effect of offering result returns on consent among our face-to-face sample (n=2286). Similarly, result return is offered to half of our web sample, which totals 3,824 participants. The study aims to provide insights into whether offering to return personalized reports increases consent and whether the effect is different by mode.
Dr Colter Mitchell (University of Michigan) - Presenting Author
In the past 30 years, structural MRI and functional MRI technologies (typically accompted with surveys) have advanced, allowing scientists to study the brain's influence on behavior and health. For research to be useful, it must externally valid, which is challenging due to the complexity of brain data collection methods compared to most other biomarker collections. Current efforts to improve neuroimaging research involves focusing on sample size, sharing reproducible code, and pre-registering hypotheses. However, reporting practices across demographics and methods, and reliance on non-probability recruitment techniques suggest it is far from generalizable despite relatively larger samples. We find that larger sample sizes were associated with higher citation rates but not with better reporting practices and limited linkages to generalizability. We utilize multiple samples to examine predictors of participation in neuroimaging studies. The Study of Adolescent to Adult Neurodevelopment (SAND; n=606) is a subsample of the Future of Families and Child Wellbeing study (which is representative of births in large US cities from 1998-2000). Overall, around 51% of FFCW study members were able to also participate in SAND. However, an additional 22% of FFCW study members would have participated but were excluded due to common MRI protocols (e.g tattoos, high BMI, etc). Many of these factors have strong social gradients and thus impact the generalizability of the data unless accounted for post-hoc. Also, 98% of SAND participants were willing to provide other biomarkers (anthropometric, blood draw, saliva). This suggests that the MRI is a far more difficult biomarker collection compared to many other biomarkers used in survey research. The primary variables related to willingness to participate were family income/SES (where a U-shaped relationship is present), race, and participation in more waves of FFCW. We also examine who is willing to participate in two
Dr Tugba Adali (Centre for Longitudinal Studies, UCL Social Research Institute, University College London)
Mr Matt Brown (Centre for Longitudinal Studies, UCL Social Research Institute, University College London)
Dr Alessandra Gaia (Centre for Longitudinal Studies, UCL Social Research Institute, University College London) - Presenting Author
Professor Morag Henderson (Centre for Longitudinal Studies, UCL Social Research Institute, University College London)
Dr Darina Peycheva (UCL Institute of Epidemiology & Health, formerly Centre for Longitudinal Studies, UCL Social Research Institute, University College London)
Dr Tim Morris (Centre for Longitudinal Studies, UCL Social Research Institute, University College London)
Social surveys now often collect biological samples in order to obtain objective measures of health. Increasingly, DNA is extracted to facilitate analysis of the interplay between genetics and social behaviour. The research utility of biological samples is enormous, but the collection can be extremely challenging, and increasing use of web and other remote methods for conducting surveys makes these challenges even more pronounced.
This presentation describes the collection of saliva samples for genotyping in the 9th wave of a large-scale longitudinal study of young adults in England: the Next Steps Age 32 survey. Prior waves of data collection have primarily focused on socio-economic factors and participants have never previously been asked to provide any form of biomeasure of biological sample. The request to provide a saliva sample therefore marked a significant transition for the study. The study uses a web-first mixed mode design so a further challenge was motivating web participants to consent and then post back their samples. We will describe the protocol, the consent rates, return rates and the quality of the samples for genotyping. We will analyse the factors associated with the likelihood of consenting and returning a sample (including the impact of an increase in incentives during fieldwork) and will describe the reasons given by those unwilling to consent. We conclude with recommendations for survey practice including suggestions for collection protocols and respondent materials.
Mrs Hayley Cheshire (Inuvi Health) - Presenting Author
Mr Timothy Woolley (Inuvi Diagnostics)
The inclusion of blood sampling remains a key priority for many social surveys due to the vast array of laboratory tests available to social scientists. Whilst trained phlebotomists have traditionally been used to collect ‘gold-standard’ venous samples from study participants, there is increasing interest in how capillary blood samples can be collected by participants themselves using self-administered devices, and whether this presents a valid alternative to venous sample collection. Use of self-administered devices helps to reduce survey costs associated with blood collection and provides a less invasive, less painful, and more ‘patient-centric’ approach (Hoffman et al., 2023). However, there remains uncertainty amongst the research community around what capillary sampling devices are suitable for use with specific populations and the reliability of the blood results that are generated (Boffel et al., 2024).
For this presentation we will draw upon the experience of our own company, Inuvi Diagnostics, in the use of lancet-based devices to collect finger-prick capillary blood, as well as exploring other capillary blood devices currently available. We will review a range of public resources to evaluate different types of devices in relation to their utility within the social survey context - e.g. the participant experience, transportation logistics, and suitability of use. In addition, we will present our own laboratory validation data used for UKAS accreditation to demonstrate how capillary and venous blood results are comparable across a range of blood analytes commonly tested in social surveys. We will provide an overview of which blood analytes can be reliably measured using capillary samples, and present supplementary laboratory data to show the stability of capillary samples over time, the impact of temperature, and haemolysis rates for specific analytes.
These findings will help inform researchers about the appropriate use of capillary blood sampling in social surveys.