"Blood protein levels predict leading incident diseases and mortality in UK Biobank"
https://www.medrxiv.org/content/10.1101/2023.05.01.23288879v...
( May 03, 2023 ; open, pre-print; not-yet peer-reviewed ; Alzheimer’s dementia included .. )
Abstract:
"The circulating proteome offers insights into the biological pathways that underlie disease. Here, we test relationships between 1,468 Olink protein levels and the incidence of 23 age-related diseases and mortality, ascertained over 16 years of electronic health linkage in the UK Biobank (N=49,234). We report 3,123 associations between 1,052 protein levels and incident diseases (PBonferroni < 5.4×10−6). Forty-four proteins are indicators of eight or more morbidities. Next, protein-based scores (ProteinScores) are developed using penalised Cox regression. When applied to test sets, eight ProteinScores improve Area Under the Curve (AUC) estimates for the 10-year onset of incident outcomes (PBonferroni < 0.0025) beyond age, sex and additional health and lifestyle covariates. The type 2 diabetes ProteinScore outperforms HbA1c (P = 5.7×10−12) – a clinical marker used to monitor and diagnose type 2 diabetes. A maximal type 2 diabetes model including the ProteinScore, HbA1c and a polygenic risk score has AUC = 0.90 and Precision-Recall AUC = 0.76. These data characterise early proteomic contributions to major age-related disease."
Alzheimer dementia * 10 proteins:
(jpg) https://www.medrxiv.org/content/medrxiv/early/2023/05/03/202...
And you can check the data ( protein * disease ) :
"Our Shiny https://protein-disease-ukb.optima-health.technology app [Username: ukb_disease, Password: shinyappUKB] provides visualisations for sensitivity analyses run across cases over successive years of follow up, allowing for interrogation of individual protein-outcome relationships."