A blood test for pregnancy-related metabolic changes could replace the stigmatising and inaccurate use of BMI to predict which women will develop gestational diabetes. Current NHS guidelines use a BMI of 30 or above as the sole trigger for gestational diabetes screening, but this misses 51% of cases—women who develop the condition despite having a BMI below 30. BMI also fails to distinguish fat from lean mass and performs poorly across ethnic groups. The SHAPES-Bio project will test whether adding five specific blood biomarkers—including adiponectin, sex-hormone binding globulin, and ferritin—to body fat measurements can push risk prediction accuracy from the current "fair" range into something clinically useful. If successful, the research could replace a one-size-fits-all BMI threshold with personalised risk triage for the roughly 185,000 pregnancies per year in England and Wales currently classified as high-risk. This would reduce unnecessary interventions for low-risk women while catching more cases of gestational diabetes early, cutting NHS resource waste and improving long-term health outcomes for both mothers and children. The project also maps existing datasets for future validation, laying groundwork for a larger implementation study.
View original technical description
Research question: Does adding biomarkers to adiposity risk-prediction model(s) improve accuracy in predicting pregnancy complications? Background: Maternal obesity increases risk of pregnancy complications, including gestational diabetes (GDM), maternal/perinatal mortality, and longer-term obesity and type 2 diabetes for women and children. Obesity is usually defined as BMI≥30kg/m2. With rising obesity rates, ~185,000 pregnancies/year in England and Wales require "high-risk" obstetric care. While BMI is routinely used to stratify risk and triage care, it is a poor predictor of individual risk, particularly among women and some ethnic groups, as it doesn t distinguish between fat and lean mass. Qualitative studies and SHAPES PPIE members describe that BMI is stigmatising, inaccurately classifies their health status, and want more accurate measures to inform pregnancy care. Measures of body fat amount and distribution (adiposity) may work better than BMI and be more acceptable to pregnant women/people. Preliminary results of our ongoing SHAPES project show that using BMI≥30kg/m2 as an independent risk factor for GDM screening (current guidelines) only identified 49% of women who developed GDM (51% had a BMI<30kg/m2 and no other risk factor). BMI≥30.0kg/m2 performed poorly for risk prediction (AUROC:0.657). This improved when BMI was combined with waist:height (AUROC:0.714) or ultrasound abdominal visceral fat measures (AUROC:0.716). However, gains were marginal, and only in the "fair" range of risk prediction accuracy. Biomarkers offer potential for enhancing risk prediction due to pregnancy-induced metabolic changes that BMI/adiposity measures cannot capture alone. Aims: To inform triage of care to improve maternal and infant health by 1) determining whether adding biomarkers to adiposity risk prediction model(s) for pregnancy complications can improve accuracy compared to current practice 2) identifying existing datasets for future external validation through individual participant data (IPD) meta-analysis Methods/work packages (WP): WP1. SHAPES-Bio Risk Prediction Study: Data from the SHAPES study will be used, with serum samples from 962 participants. Biochemical measures will include a traditional lipid panel, adiponectin, gamma-glutamyl transferase, sex-hormone binding globulin, and ferritin. Robust predictive models will be built allowing for missing data and using resampling methods to calculate optimism corrected performance. Decision curve analysis will compare the clinical net benefit of each model with current practice. WP2. Scoping Review: Risk prediction models must be validated before implementation. SHAPES uses IPD meta-analysis methods for external validation of the cohort results in heterogeneous existing datasets. SHAPES-Bio will build on this, employing systematic search methods following PRISMA-ScR guidelines to identify studies with data on adiposity measures, biomarkers, and pregnancy outcomes. Data will be charted to map existing datasets suitable for external validation in future research. Dissemination/impact: The findings will be shared with pregnant women, the public, health professionals, maternity services, policymakers, and researchers through newsletters, project website, policy briefs, social media, academic journals, and conferences. PPIE co-development of outputs will be conducted. Should promising model(s) be identified, external validation, economic evaluation and implementation research of SHAPES-Bio will be included in a future funding application. The anticipated impact includes improved risk-prediction models, reduced pregnancy complications, decreased NHS resource waste, and more personalised triage of pregnancy care. Timeline: 01/08/2025-31/10/2026
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know