Daily Endocrinology Research Analysis
Analyzed 61 papers and selected 3 impactful papers.
Summary
Three impactful endocrinology studies span basic mechanisms, AI-enabled epidemiology, and drug safety. Iron sufficiency was identified as a maturation-critical cue for pancreatic β-cells, AI-predicted insulin resistance associated with elevated risks across multiple cancers in UK Biobank, and GLP-1 receptor agonists were linked to a higher short-term risk of NAION versus DPP-4 inhibitors.
Research Themes
- Endocrine–metabolic mechanisms driving β-cell maturation
- AI-enabled risk stratification linking insulin resistance to cancer
- Drug safety signals for GLP-1 receptor agonists (ocular ischemic neuropathy)
Selected Articles
1. Iron deficiency induces maturation-dependent loss of pancreatic β-cells.
This mechanistic study shows that iron sufficiency is critical specifically during the β-cell maturation window. Iron restriction via chelation or TFRC disruption impairs oxidative metabolism and survival of immature but not mature β-cells, revealing a developmental switch in iron dependency.
Impact: It uncovers iron as a key metabolic cue governing β-cell maturation, providing actionable insight for generating fully functional stem cell-derived β-cells and reframing pediatric iron deficiency as a potential risk for impaired islet development.
Clinical Implications: While preclinical, the findings suggest considering iron sufficiency in perinatal and pediatric nutrition for islet health and optimizing iron handling within protocols to mature stem cell-derived β-cells for transplantation.
Key Findings
- Iron is essential during the β-cell maturation transition, but mature β-cells are relatively resilient to iron depletion.
- Chemical chelation and genetic disruption of TFRC-mediated iron uptake impair oxidative metabolism and survival in immature β-cells.
- Identifies a developmental switch in iron dependency, informing strategies to mature stem cell-derived β-cells.
Methodological Strengths
- Convergent evidence from mouse and human β-cell systems with both pharmacologic and genetic perturbations.
- Direct assessment of metabolic function and cell survival across developmental stages.
Limitations
- Preclinical study without direct clinical outcome data.
- Timing and reversibility of iron restriction effects in vivo across human developmental windows remain to be defined.
Future Directions: Define precise developmental windows for iron supplementation, test reversibility in vivo, and incorporate iron handling into protocols for differentiating and maturing stem cell-derived β-cells.
Pancreatic β-cells maintain glucose homeostasis by secreting insulin in response to rising blood glucose, a process fueled by mitochondrial ATP production. Iron, a core cofactor in the electron transport chain, is essential for this metabolic coupling. While the cytotoxic effects of iron overload are well known, the role of iron sufficiency during β-cell development remains unclear. Here, we identify a maturation-dependent requirement for iron in mouse and human β-cells. Using chemical chelation and genetic disruption of transferrin receptor (TFRC)-mediated uptake, we show that immature β-cells depend on iron during metabolic transition to functional maturity. Iron restriction at this stage impairs oxidative metabolism and compromises survival. In contrast, mature β-cells remain resilient to iron depletion, revealing a developmental switch in iron dependency. These findings establish iron as a key metabolic cue in β-cell development and suggest strategies to generate fully functional stem cell-derived β-cells for diabetes modeling and cell replacement therapy.
2. Machine learning-predicted insulin resistance is a risk factor for 12 types of cancer.
An ML-derived insulin resistance score (AI-IR) outperformed traditional indices for predicting diabetes and was associated with higher risks of multiple cancers in the UK Biobank. The results implicate insulin resistance as a shared upstream driver of oncogenesis across diverse tissues.
Impact: Establishes a scalable, clinically accessible ML proxy of insulin resistance that links metabolic dysfunction to cancer risk across multiple sites, opening avenues for prevention-focused risk stratification.
Clinical Implications: AI-IR could help identify individuals who may benefit from intensified lifestyle/metabolic interventions and tailored cancer surveillance; however, external validation and clinical utility trials are needed before changing screening guidelines.
Key Findings
- AI-IR outperformed BMI, MetS, TG/HDL ratio, and TyG index for predicting incident diabetes in UK Biobank.
- AI-IR was significantly associated with increased risks of uterine, kidney, esophageal, pancreatic, colon, and breast cancers, with nominal associations for six additional cancer types.
- Composite cancer risk increased with AI-IR (age- and sex-adjusted HR 1.25, 95% CI 1.20–1.31).
Methodological Strengths
- Large-scale cohort (UK Biobank) with standardized phenotyping and robust modeling.
- Direct comparison of AI-IR with multiple conventional insulin resistance surrogates.
Limitations
- Observational design limits causal inference and residual confounding may persist.
- External validation beyond UK Biobank and assessment of clinical utility are required; potential detection/surveillance biases were not fully addressed.
Future Directions: Externally validate AI-IR across diverse populations, test risk-adapted prevention/screening strategies, and dissect biological mediators linking insulin resistance to site-specific carcinogenesis.
Insulin resistance is suggested to be a risk factor for cancer; however, large-scale epidemiological evidence linking insulin resistance to cancer remains limited. Here we apply a machine learning-based prediction model of insulin resistance with nine clinical parameters, termed artificial intelligence-derived insulin resistance (AI-IR), to the UK Biobank and demonstrated that AI-IR exhibits the highest predictive performance for diabetes incidence compared to body mass index (BMI), metabolic syndrome (MetS), triglyceride to high-density lipoprotein cholesterol (TG/HDL) ratio, and triglyceride-glucose (TyG) index. Moreover, AI-IR is significantly associated with an increased risk of six cancers (uterine, kidney, esophagus, pancreas, colon, and breast) and showed nominal associations with six additional cancers (renal pelvis, small intestine, stomach, liver and gallbladder, leukemia, and bronchial and lung). When we define composite cancers by merging cancer types whose risks increase with AI-IR, age- and sex-adjusted hazard ratio is 1.25 (95% confidence interval, 1.20-1.31; P < 1 ×10
3. Glucagon-Like Peptide 1 Receptor Agonists and Risk of Nonarteritic Anterior Ischemic Optic Neuropathy in Patients With Type 2 Diabetes.
In a large UK database emulating a target trial, initiating GLP-1 RAs was associated with a higher 1-year NAION risk than DPP-4 inhibitors (RR 2.56), most pronounced within the first 6 months and in younger individuals, men, ever-smokers, and those with ≥1% HbA1c reduction.
Impact: Given widespread GLP-1 RA use, this safety signal for NAION is clinically consequential and time-sensitive, informing patient counseling and early symptom surveillance.
Clinical Implications: Clinicians should counsel GLP-1 RA initiators—especially younger men, ever-smokers, and those with rapid HbA1c reductions—about NAION symptoms (sudden painless vision loss) and consider prompt ophthalmologic evaluation if suspected. Risk–benefit remains favorable for most, but vigilance is warranted in the first 6 months.
Key Findings
- At 1 year, NAION incidence was 18.5/100,000 in GLP-1 RA initiators vs 7.2/100,000 in DPP-4 inhibitor initiators; RR 2.56 (95% CI 1.44–4.86).
- Risk was highest in the first 6 months and in younger patients (<50 years), men, ever-smokers, and those with ≥1% HbA1c reduction.
- Absolute risk difference was 11.3 per 100,000 over 1 year.
Methodological Strengths
- Active-comparator, new-user design with target trial emulation and propensity-score fine-stratification weighting.
- Very large sample with prespecified safety outcome and subgroup analyses by duration and risk factors.
Limitations
- Observational database study subject to residual confounding and outcome misclassification.
- Lack of ophthalmologic adjudication and potential channeling bias cannot be fully excluded.
Future Directions: Independent replication with ophthalmologic adjudication, mechanistic studies on optic nerve perfusion and glycemic dynamics, and risk-mitigation strategies (e.g., gradual HbA1c reduction) should be evaluated.
OBJECTIVE: To estimate the effect of initiating glucagon-like peptide 1 receptor agonists (GLP-1 RAs) versus dipeptidyl peptidase 4 (DPP-4) inhibitors on incident nonarteritic anterior ischemic optic neuropathy (NAION) among adults with type 2 diabetes. RESEARCH DESIGN AND METHODS: This active-comparator, new-user cohort emulated a pragmatic target trial using the U.K. Clinical Practice Research Datalink. Patients aged ≥18 years with a physician diagnosis of type 2 diabetes who newly initiated a GLP-1 RA or a DPP-4 inhibitor were included. DPP-4 inhibitors were selected as the comparator because they may be used as second-line treatment, like GLP-1 RAs, and have no established association with NAION. We estimated risks, risk differences (RDs), and risk ratios (RRs) of incident NAION, adjusted using propensity-score fine-stratification weighting. RESULTS: At 1 year, there were 14 NAION events among 106,858 GLP-1 RA initiators (18.5 per 100,000) and 53 among 416,369 DPP-4 inhibitor initiators (7.2 per 100,000). GLP-1 RAs were associated with an increased risk of NAION compared with DPP-4 inhibitors (RR 2.56; 95% CI 1.44-4.86; RD 11.3 per 100,000). The risk of NAION was higher during the first 6 months of use, diminishing with longer duration of use, and was higher in patients aged <50 years, men, ever-smokers, and those with ≥1% hemoglobin A1c reduction. CONCLUSIONS: GLP-1 RAs were associated with an increased 1-year risk of NAION compared with DPP-4 inhibitors among adults with type 2 diabetes, particularly in younger patients, men, ever-smokers, and patients with a marked hemoglobin A1c reduction.