Tracing the Genetic Footprint of the UK National Health System

Speaker: Dr Nicolau Martin Bassols

Affiliation: University of Bologna

Location: Room 104, Chamberlain Building (#35), St Lucia Campus

Zoom: https://uqz.zoom.us/j/82603079317

Abstract: Estimates of the impact of early-life events on surviving individuals may be distorted by selective survival or mortality, which is often difficult to detect. This paper introduces a method for identifying and quantifying this type of bias using genetic data. In the absence of selection, Polygenic Indices (PGIs), predetermined at conception, should be equally distributed across treatment and control groups. Systematic differences in PGIs following a shock or policy therefore indicate selective survival.  We apply this approach to the introduction of the UK National Health Service (NHS) in July 1948, using data from the UK Biobank and a regression discontinuity design under local randomization. We find that the NHS reduced infant mortality by 6.5 per 1,000 births—a decline of 18.5\%—likely due to increased hospital births. This mortality decline altered the genetic composition of surviving cohorts: NHS exposure increased PGIs for adverse conditions (e.g., depression, chronic obstructive pulmonary disease) and decreased PGIs for positive traits (e.g., educational attainment, self-rated good health). Results hold true in sibling fixed effect designs, different survey data, and on PGIs directly linked to early-life conditions. The effects are primarily driven by the survival of individuals born in the most socioeconomically disadvantaged areas. We identify a significant survival bias and highlight the novel use of genetic endowment data to quantify it, thereby enabling more valid and policy-relevant inference.