Abstract Body (Do not enter title and authors here): Background: High blood pressure (BP) is a leading cause of cardiovascular disease (CVD), however, the biological mechanisms underlying its development are incompletely understood.
Hypothesis: We hypothesized that longitudinal profiling of the plasma proteome integrated with human genetics would identify novel determinants of BP and CVD.
Methods: Plasma proteins (n=2,886) were measured using a proximity extension assay in 2,008 adults from the Multi-Ethnic Study of Atherosclerosis at 3 different visits spanning 15 years of follow-up. Systolic BP (SBP) was measured at each timepoint and adjusted for BP medication via fixed addition. We evaluated the association between protein abundance and SBP both cross-sectionally and through time-updating linear mixed models, with subsequent validation in the Jackson Heart Study (JHS) (n=1,055). Potential causal roles of SBP-associated proteins were explored through cis-Mendelian Randomization (cis-MR) and colocalization experiments. SBP-related proteins were further tested for associations between their levels or trajectories with incident CVD through cox proportional hazards and joint models respectively.
Results: Serial profiling identified 5-fold more proteins associated with SBP than a cross-sectional analysis alone (725 vs 125 at FDR<5%) with similarly improved validation in JHS (256 vs 48). SBP-associated proteins included renin, NT-proBNP, and novel circulating markers such as protein tyrosine phosphatase receptor B (β=2.79, p=8.03e-22). cis-MR supported potential causal roles for 12 proteins in BP regulation, of which only 2 would have been identified in a cross-sectional analysis (Table 1). Further, we found strong evidence for shared genetic control between plasma natural killer cell cytotoxicity receptor 3 ligand 1 and SBP (H4 PP >99.99%). Finally, levels of 40 proteins were associated with incident CVD though associations between protein trajectories and CVD were attenuated after adjustment for baseline abundance.
Conclusions: Serial profiling of the plasma proteome improves power for detecting markers of BP and CVD and uncovers novel pathways of BP regulation, highlighting the potential of longitudinal molecular data for both biological and biomarker discovery.