A recent Endocrinology digest,
as sent to subscribers

Study Type: Erratum / Correction notice

Key Question: Not applicable — this notice corrects an omission of funding acknowledgement in a published RCT examining an eating disorder prevention programme for young women with Type 1 Diabetes.

Key Findings:

• No new data or findings are presented
• The sole correction adds a previously omitted NIH grant acknowledgement (P30 DK036836, Joslin Diabetes Center Clinical Research Center)
• The original article (Diabetes Care 2025;48:220–225) remains otherwise unchanged

Clinical Relevance: The underlying RCT addresses disordered eating in Type 1 Diabetes — a clinically significant comorbidity relevant to UK endocrinology and diabetes services — but this notice contains no new clinical information.

Limitations: Not applicable — this is an administrative correction only; clinicians should refer to the original article (DOI: 10.2337/dc24-1599) for study content.

Study Type: Narrative review / commentary

Key Question: What is the current state of hybrid automated insulin delivery (AID) systems, and what developments are needed to achieve fully automated next-generation systems?

Key Findings:

• Hybrid AID systems have successfully moved from research to commercial availability, with increasing choice influencing system selection by both users and clinicians
• Next-generation fully automated approaches under development include ultrarapid insulins, adjunctive therapies (e.g., pramlintide, GLP-1 RAs), additional wearable signal integration, and dual-hormone systems
• Emerging evidence supports AID use beyond typical type 1 diabetes populations, including type 2 diabetes, cystic fibrosis-related diabetes, and inpatient settings

Clinical Relevance: With NHS England expanding AID commissioning under the Long Term Plan, this review contextualises system selection, equity of access, and the pipeline of technologies likely to reach UK clinical practice.

Limitations: As a narrative review, findings reflect authors' synthesis without systematic methodology or formal quality assessment.

Study Type: Prospective cohort study with internal validation and external replication

Key Question: Can plasma proteomic signatures improve prediction of 3-year progression from impaired fasting glucose to type 2 diabetes?

Key Findings:

• Six proteins were independently associated with IFG-to-diabetes progression: lower PTPRS, ANTXR2, ADIPOQ, CNTFR, TMEM132C, and higher ADAMTSL2; implicated pathways included carbohydrate metabolism and glycolysis
• Adding these six proteins to conventional risk factors improved discrimination (optimism-corrected AUC 0.81, ΔAUC +0.03, p=0.005) and net reclassification (~12%) over clinical variables alone
• Two of six proteins (ADIPOQ, ADAMTSL2) replicated in the independent multi-ethnic MESA cohort

Clinical Relevance: These proteomic markers could refine diabetes risk stratification in people with IFG, potentially informing more targeted prevention interventions — relevant to NHS prediabetes pathways including the National Diabetes Prevention Programme.

Limitations: Proteomics platforms are not yet clinically deployable at scale, limiting immediate NHS applicability; only two proteins replicated externally.

Study Type: Randomised, double-blind, placebo-controlled crossover trial

Key Question: Does exogenous melatonin impair glucose homeostasis differently in carriers versus non-carriers of the MTNR1B type 2 diabetes risk variant?

Key Findings:

• In MTNR1B G-allele carriers, melatonin worsened glucose tolerance (+11.7%, 95% CI 1.0–22.3) and suppressed first-phase β-cell responsivity by 40% (95% CI −52.4 to −24.3; p=0.0003) compared with placebo; no significant effect in non-carriers
• Melatonin slowed insulin-induced suppression of second-phase insulin secretion in carriers (+64.3%, 95% CI 23.1–119.2; p=0.001), suggesting disrupted insulin negative feedback
• Melatonin prevented exogenous insulin-induced hypoglycaemia exclusively in carriers (0 vs 7 events; p=0.001)

Clinical Relevance: Given the sharp rise in over-the-counter melatonin use in the UK, clinicians should consider MTNR1B genotype status when advising patients with or at risk of type 2 diabetes about melatonin supplementation.

Limitations: Small sample (n=21), restricted to healthy European-ancestry adults, limiting generalisability to clinical populations.

Study Type: Retrospective analysis with tool validation

Key Question: Can an automated EMR search tool (PEPPER) accurately and efficiently identify patients with A-β+ ketosis-prone diabetes (KPD) within a large paediatric type 2 diabetes dataset?

Key Findings:

• PEPPER reviewed 1,660 charts and identified 110 with T2D plus DKA within 6 months of diagnosis; 21 met full A-β+ KPD criteria
• Review time was significantly reduced versus manual searching (13.4 ± 3.9 s vs. 26.6 ± 9.4 s per chart; P <0.001)
• PEPPER achieved 100% accuracy against manual review as the reference standard

Clinical Relevance: KPD is likely underdiagnosed in UK minority ethnic populations; scalable EMR screening tools could support systematic identification within NHS diabetes registers.

Limitations: Validated in a single-centre paediatric cohort only, limiting generalisability to adult populations and different EMR systems used across NHS trusts.

Study Type: Prospective randomised crossover trial

Key Question: Does real-time CGM improve glycaemic outcomes compared with capillary blood glucose monitoring in adults with type 2 diabetes on maintenance haemodialysis?

Key Findings:

• Primary outcome (time below range <70 mg/dL) did not differ significantly between rtCGM and CBG groups (1.17% vs 1.29%; p=0.28)
• rtCGM was associated with significantly higher time in range (63.4% vs 54.5%) and lower mean glucose (173.6 vs 187.7 mg/dL; both p≤0.01)
• Time above range >180 mg/dL (35.3% vs 44.3%) and >250 mg/dL (12.3% vs 18.8%) were meaningfully reduced with rtCGM (both p≤0.01)

Clinical Relevance: Supports consideration of rtCGM in haemodialysis patients with type 2 diabetes, a group at high hypoglycaemia risk where HbA1c is unreliable — relevant to emerging NHS CGM prescribing guidance.

Limitations: Crossover design with a relatively small sample limits generalisability and statistical power.

Study Type: Cohort study (longitudinal and cross-sectional analyses)

Key Question: Do polygenic scores (PS) for macronutrient intake predict dietary behaviours and cardiometabolic outcomes across childhood and adolescence?

Key Findings:

• Higher carbohydrate PS was associated with increased odds of consuming ≥2 weekly servings of sugar-sweetened beverages (OR 1.20; 95% CI 1.07–1.33) and ≥1 weekly fast-food meal (OR 1.11; 95% CI 1.004–1.23) from ages 3–18 years.
• Higher protein PS was associated with reduced odds of SSB consumption (OR 0.84; 95% CI 0.75–0.93).
• No consistent associations were identified between macronutrient PS and cardiometabolic outcomes at age 18.

Clinical Relevance: Identifies a genetic basis for appetite-driven dietary patterns in youth, potentially informing personalised approaches to obesity and metabolic risk prevention in paediatric endocrinology.

Limitations: Sample restricted to non-Hispanic White children from a single US cohort, substantially limiting generalisability to NHS patient populations.

Study Type: Retrospective policy analysis using secondary public data

Key Question: What is the reach of US state-level insulin out-of-pocket (OOP) cost cap legislation among commercially insured adults with diabetes?

Key Findings:

• By 2026, 29 states plus DC had enacted insulin OOP caps, covering approximately 990,000 eligible adults on insulin in state-regulated plans
• 670,000 adults in states without caps and a further 2.2 million in federally regulated plans (e.g., self-insured employer schemes) remain entirely unprotected by state legislation
• The majority of commercially insured insulin users fall outside state policy jurisdiction

Clinical Relevance: While directly a US policy issue, this highlights the structural limitations of regionally fragmented insulin affordability schemes — relevant to NHS England discussions around formulary access, biosimilar insulin uptake, and health inequity in diabetes management.

Limitations: Estimates rely on aggregated public data rather than individual-level insurance or prescribing records, limiting precision.