Collecting saliva samples for DNA genotyping in a large-scale cohort study of young adults in England

Social surveys are increasingly incorporating DNA collection to unlock research opportunities in social and medical sciences. Saliva sampling is emerging as a common and practical method - particularly well-suited to self-completion surveys where interviewers or other trained professionals are not present. However, a number of methodological challenges in the collection of saliva samples remain unsolved. We analyse data from a cohort study of young adults in England: the Next Steps Age 32 survey to explore these methodological challenges. Approximately 57 per cent of eligible cohort members consented to provide a saliva sample and 27 per cent returned a sample to the laboratory, resulting in genotyped data for 24 per cent. Both consent and sample return were significantly associated with participant characteristics - such as white ethnicity - as well as survey-related factors, including engagement with study materials, prior wave participation and consent to health data linkage. Experimental evidence also indicates that higher monetary incentives (£10 versus £5) increased both consent rates (57.8 per cent versus 53 per cent) and sample returns (29.8 per cent versus 20.6 per cent). Analysis of non-consent reasons revealed that many responses were vague, while 26.9 per cent mentioned they were uncomfortable with the task, found it intrusive or expressed privacy concerns. We conclude with recommendations for improving biosample collection in survey practice. Our work contributes to the growing literature on integrating biomarker collection into large-scale, mixed-mode, multi-purpose social research.