A transcriptomic signature that predicts prehypertension in adolescence and higher systolic blood pressure in childhood
Reena Perchard, Terence Garner, Philip G. Murray, Amirul Roslan, Lucy E. Higgins, Edward D. Johnstone, Adam Stevens, Peter E. Clayton
Reena Perchard, Terence Garner, Philip G. Murray, Amirul Roslan, Lucy E. Higgins, Edward D. Johnstone, Adam Stevens, Peter E. Clayton
View: Text | PDF
Clinical Research and Public Health Cardiology Public Health

A transcriptomic signature that predicts prehypertension in adolescence and higher systolic blood pressure in childhood

Abstract

BACKGROUND Suboptimal fetal growth (SFG), being born small for gestational age (SGA), and catch-up (CU) growth are, individually and together, linked to cardiometabolic risks. However, not all develop adverse outcomes. This study aimed to validate a transcriptomic signature to identify individuals at greatest cardiometabolic risk.METHODS Using National Heart, Lung and Blood Institute (NHLBI) criteria to define cardiometabolic risk, healthy and prehypertensive 17-year-olds were identified in the Avon Longitudinal Study of Parents and Children (ALSPAC) (UK) childhood cohort. Epigenomic and transcriptomic differences were analyzed. A hypergraph identified functionally related genes, which were used in random forest classification to predict prehypertensive phenotypes. The BabyGRO (UK) cohort included 80 children aged 3–7 years, born at term following pregnancies with SFG risks. Anthropometric and cardiometabolic markers and transcriptomic profiles were collected, fetal and childhood weight trajectories and their relationship to cardiometabolic markers were assessed, and transcriptome was used for prediction.RESULTS Individuals with CU-SGA in ALSPAC were 1.6 times more likely than all others to be prehypertensive at 17 years (P < 1 × 10–5). A 42-gene hypergraph cluster was highly predictive of prehypertension (AUC 0.984, error rate 5.4%). In BabyGRO, 20 of these genes accurately predicted higher systolic blood pressure (AUC 0.971, error rate 3.6%). This transcriptomic signature could help identify children with adverse pre- and postnatal growth who may develop prehypertension.CONCLUSION A blood transcriptomic signature exists in childhood which distinguishes those at risk of adult cardiometabolic disease among children with adverse pre- and postnatal growth.TRIAL REGISTRATION Regional ethics committee reference 17/NW/0153, IRAS project ID 187679.FUNDING Centre grant to the Maternal and Fetal Health Research Centre by Tommy’s The Pregnancy and Baby Charity, Child Growth Foundation, European Research Council funding as part of the Health and Environment-wide Associations based on Large Population Surveys (HEALS) study

Authors

Reena Perchard, Terence Garner, Philip G. Murray, Amirul Roslan, Lucy E. Higgins, Edward D. Johnstone, Adam Stevens, Peter E. Clayton

×

Figure 8

Hypergraphs.

Options: View larger image (or click on image) Download as PowerPoint
Hypergraphs. In a traditional network, an edge (represent here by the li...
In a traditional network, an edge (represent here by the line) connects 2 nodes together. In a hypergraph, an edge can represent a relationship between any number of nodes — e.g., multiple proteins interacting to form a protein complex. Not only can edges connect any number of nodes, but the same pair of nodes can be connected by multiple edges. Altogether, this allows us to increase the depth of the information we can capture. The hypergraph central clusters represent subsets of significant genes and metabolites sharing the most correlations and indicating likely functional relationships.