Publication type
Understanding Society Working Paper Series
Series Number
2023-05
Series
Understanding Society Working Paper Series
Authors
Publication date
March 14, 2023
Abstract:
Proteomic methods have been developed to enable measurements of low levels of proteins in blood samples. In Waves 2 and 3 (2010-2012) of Understanding Society, blood samples were collected in participants’ homes, then posted to the laboratory and subsequently used for the measurement of 184 proteins. It is unknown whether the delay between blood collection and laboratory processing compromises blood sample integrity such that genuine associations between proteins and demographic factors are obscured. This paper describes the protein measurements and the quality checks performed on them. It also describes the proteins’ associations with a range of factors, including pre-analytic factors (time of day of sample collection, transit delay: number of days the sample was in the post, haemolysis of sample) and demographic factors (age, sex, educational attainment, geographical region, ethnicity). A range of standard statistical tests were used. We find that protein measurements in the dataset are highly reproducible, with low levels of missingness. We observe low detectability in around one-tenth of proteins. We identify transit delay as an important source of variation, and we characterise the nature of its relationship to each individual protein. Examining educational attainment as an exemplar, we observe that of the proteins measured, 127 varied by educational attainment, including just under half of the proteins defined with low detectability. The largest associations with educational attainment are found in proteins previously described to be associated with cognitive function. While adjustment for pre-analytic factors usually improved overall model fit, association of proteins with educational attainment were largely unaltered.
Subjects
Paper download#547688