Daily Endocrinology Research Analysis
Three impactful endocrinology studies advance screening and risk stratification. A new DiabetesLiver score accurately identifies advanced liver fibrosis and predicts liver outcomes in type 2 diabetes. National data from Türkiye show how optimizing second‑tier markers can markedly improve congenital adrenal hyperplasia screening performance, while a multicenter Bangladesh study proposes a lower random capillary glucose cut-off to diagnose diabetes in resource-limited settings.
Summary
Three impactful endocrinology studies advance screening and risk stratification. A new DiabetesLiver score accurately identifies advanced liver fibrosis and predicts liver outcomes in type 2 diabetes. National data from Türkiye show how optimizing second‑tier markers can markedly improve congenital adrenal hyperplasia screening performance, while a multicenter Bangladesh study proposes a lower random capillary glucose cut-off to diagnose diabetes in resource-limited settings.
Research Themes
- Non-invasive risk algorithms for liver fibrosis in type 2 diabetes
- Optimization of endocrine newborn screening using multiplex steroid markers
- Pragmatic diabetes diagnostics in resource-limited settings
Selected Articles
1. DiabetesLiver score: A non-invasive algorithm for advanced liver fibrosis and liver-related outcomes in type 2 diabetes mellitus population.
Across multi-cohort development and validation, the DiabetesLiver score (waist circumference, ALT, AST, platelets, albumin) accurately identified advanced fibrosis (AUC 0.835–0.870; external 0.823) and predicted liver-related outcomes. Dual cut-offs (2.39 and 3.99) achieved ≥90% sensitivity or specificity and stratified T2D patients into risk tiers that tracked with future liver events in UK Biobank.
Impact: This tailored score for T2D bridges clinic and population health by enabling scalable, non-invasive fibrosis screening and prognostic stratification with external and outcome validation.
Clinical Implications: Use the DiabetesLiver score in T2D clinics to prioritize elastography/hepatology referral, guide intensity of MASLD management, and enrich high-risk groups for surveillance of liver outcomes.
Key Findings
- Developed a 5-variable score (waist circumference, ALT, AST, platelets, albumin) for advanced fibrosis in T2D.
- Performance: AUC 0.835 (derivation), 0.870 (internal validation), 0.823 (NHANES external validation).
- Dual cut-offs (2.39 and 3.99) yielded ≥90% sensitivity or specificity, enabling low/middle/high risk stratification.
- High-risk group showed increased hepatocellular carcinoma and liver-related mortality in UK Biobank.
Methodological Strengths
- Large multi-cohort development with internal and external validation
- Outcome validation for HCC and liver-related mortality in an independent biobank cohort
Limitations
- Advanced fibrosis defined by liver stiffness ≥12 kPa, not biopsy-confirmed in primary cohorts
- Potential spectrum and selection biases across heterogeneous cohorts
Future Directions: Prospective implementation studies comparing the DiabetesLiver score-guided pathways vs. usual care, and calibration across ethnicities and primary care settings.
2. The First-Year Outcomes of the Nationwide Neonatal CAH Screening in Türkiye: High Rate of False Positives for 21-Hydroxylase Deficiency and a Higher Detection Rate of Non-Classical Cases.
In its first national year (1,096,069 neonates), Türkiye’s CAH program confirmed 91 classical and 22 non-classical 21-OHD cases using a two-tier protocol. Modeling of refined cut-offs (combining succinylacetone analogs are not used here—tyrosine-based indices with 17-OHP/21-DF ratios) suggests a six-fold reduction in referrals and avoidance of second-tier testing in 95% of infants without increasing false negatives in the original dataset.
Impact: This is a rare nationwide, first-year evaluation demonstrating how analytic refinements could dramatically boost PPV and efficiency in CAH screening at scale.
Clinical Implications: Programs can adopt refined cut-offs and steroid ratios to minimize unnecessary referrals and resource use while maintaining sensitivity, improving family burden and system costs.
Key Findings
- Nationwide screening of 1,096,069 neonates identified 91 classical and 22 non-classical 21-OHD cases.
- Second-tier testing occurred in 6.88% and clinic referral in 0.27% of screened infants.
- Refined cut-offs combining 17-OHP, 21-DF, cortisol, and tyrosine-based indices could raise PPV to 72% and reduce referrals six-fold without adding false negatives in the original dataset.
- Established practical first-tier FIA-17-OHP thresholds by gestational age/weight bands.
Methodological Strengths
- Nationwide population coverage with over one million screens
- Two-tier biochemical algorithm with clear referral thresholds and scenario modeling for optimization
Limitations
- Optimization analyses are based on retrospective modeling and require prospective implementation
- Potential variability in sample timing and pre-analytical factors across centers
Future Directions: Prospective rollout of refined cut-offs with PRISMA-like monitoring of sensitivity/PPV, and integration of additional second-tier markers where cost-effective.
3. Random capillary blood glucose in the diagnosis of diabetes: a cross-sectional study in Bangladesh.
In a multicenter diagnostic accuracy study (n=3200), random capillary glucose with a cut-off ≥8.7 mmol/L outperformed the conventional ≥11.1 mmol/L threshold versus FPG, 2h-PG, and HbA1c. Diagnosis did not require hyperglycaemic symptoms, and the number needed to screen was lower than FPG and the conventional threshold.
Impact: A pragmatic, low-cost diagnostic cut-off could expand reliable diabetes diagnosis in settings where fasting or OGTT testing is impractical.
Clinical Implications: Adopting ≥8.7 mmol/L as a diagnostic threshold for random capillary glucose can streamline case-finding, reduce missed diagnoses, and lower testing burdens in primary care.
Key Findings
- Random capillary glucose strongly correlated and agreed with FPG, 2h-PG, and HbA1c across 16 centers.
- A cut-off of ≥8.7 mmol/L improved sensitivity, specificity, and AUC compared with ≥11.1 mmol/L.
- Diagnosis did not require hyperglycaemic symptoms; number needed to screen was 2.74 (lower than FPG and conventional RCBG threshold).
Methodological Strengths
- Multicenter, systematically sampled cohort with standard comparators (FPG, 2h-PG, HbA1c)
- Direct estimation of diagnostic thresholds with NNS metrics relevant to public health
Limitations
- Cross-sectional design; lacks longitudinal outcomes or repeat testing to assess stability
- Population-specific threshold may require recalibration in other regions
Future Directions: Prospective implementation studies comparing the ≥8.7 mmol/L cut-off against standard diagnostic pathways, including cost-effectiveness and impact on treatment initiation.