Daily Endocrinology Research Analysis
Analyzed 32 papers and selected 3 impactful papers.
Summary
Three endocrine-metabolic studies stand out today: a meta-analysis of randomized trials shows semaglutide consistently reduces coronary artery disease events; a population-scale EHR cohort identifies elevated non–HDL cholesterol as an independent, practical predictor of incident diabetes, hypertension, and MASLD; and real-world data from nearly 200,000 Omnipod 5 users indicate simplified meal bolusing achieves non-inferior glycemic outcomes versus carbohydrate counting. Collectively, these findings support cardio-protective incretin use, routine reporting of non–HDL-C, and broader access to automated insulin delivery without carbohydrate-counting barriers.
Research Themes
- Cardiovascular protection with incretin-based therapies
- Actionable lipid metrics beyond LDL: non–HDL-C in primary care
- Diabetes technology optimization without carbohydrate counting
Selected Articles
1. Coronary artery disease -related outcomes associated with semaglutide and tirzepatide in type 2 diabetes mellitus and obesity: a systematic review and meta-analysis.
Across 45 randomized trials, semaglutide reduced multiple CAD-related outcomes (including ACS and MI) in adults with T2DM and/or obesity, whereas evidence for tirzepatide was less precise due to fewer CAD events and limited CAD-specific reporting. Findings support semaglutide’s cardioprotection beyond glycemic and weight effects.
Impact: This meta-analysis consolidates CAD outcomes across trials and phenotypes, providing clinically actionable evidence that semaglutide reduces acute coronary events in T2DM/obesity populations.
Clinical Implications: For adults with T2DM/obesity at elevated CAD risk, semaglutide should be prioritized when selecting glucose-lowering therapy with cardioprotective potential; standardized CAD outcome reporting is needed for tirzepatide.
Key Findings
- Pooled risk reduction with semaglutide for CAD composite (RR 0.80, 95% CI 0.73-0.87)
- Lower risks observed for ACS (RR 0.77), MI (RR 0.70), UA (RR 0.82), and angina (RR 0.73)
- Tirzepatide effects were imprecise for most CAD outcomes due to fewer events and limited CAD-specific data
- Heterogeneity in CAD outcome definitions across trials underscores need for standardization
Methodological Strengths
- Systematic search across multiple databases with RCT-only inclusion
- Random-effects meta-analysis with prespecified CAD-related outcomes
Limitations
- Outcome definitions and reporting varied across trials (heterogeneity)
- Tirzepatide analyses limited by few CAD events and CAD-specific reporting
Future Directions: Conduct dedicated, adequately powered cardiovascular outcome trials for tirzepatide and standardize CAD outcome definitions/reporting across incretin trials.
AIMS: Semaglutide and tirzepatide are widely used drugs in the treatment of metabolic diseases based on incretin-based therapy. However, evidence on coronary artery disease (CAD)-related outcomes has largely been derived from separate randomized trials with different control groups, and focused synthesis across CAD phenotypes remains limited. This study aimed to evaluate CAD-related outcomes reported in randomized controlled trials (RCTs) of semaglutide and tirzepatide in adults with type 2 diabetes (T2DM) and/or obesity. DATA SYNTHESIS: We systematically searche
2. non-HDL-C: The missing parameter in primary care.
In 820,318 adults not on lipid-lowering therapy, elevated non–HDL-C was common and associated with higher baseline metabolic disease burden and increased 5-year incidence of type 2 diabetes, hypertension, and MASLD despite adjustment. Findings support routine calculation and monitoring of non–HDL-C in primary care.
Impact: This massive real-world cohort shows non–HDL-C independently predicts incident metabolic diseases beyond LDL measures, providing an immediately actionable, cost-free metric for risk stratification.
Clinical Implications: Incorporate non–HDL-C into routine lipid panels and risk discussions; use elevated non–HDL-C to intensify lifestyle counseling and consider earlier pharmacologic prevention even when LDL-C is ‘controlled’.
Key Findings
- Among 820,318 adults, 101,939 had elevated non–HDL-C by a pragmatic definition (non–HDL-C > LDL-C + 30 mg/dL)
- Elevated non–HDL-C was more prevalent in T2D, hypertension, obesity, NASH/MASLD, CKD, and CVD at baseline
- Over 5 years, elevated non–HDL-C predicted incident diabetes (HR 2.529), hypertension (HR 1.365), and MASLD (HR 1.365) after adjustment
- Propensity score matching and Cox models were used to mitigate confounding and estimate risk
Methodological Strengths
- Population-scale EHR cohort with >820,000 participants and repeated lipid measures
- Use of propensity score matching and multivariable Cox modeling
Limitations
- Retrospective design with potential residual confounding and misclassification
- Operational definition of elevated non–HDL-C may not align with all guideline thresholds
Future Directions: Prospective validation across diverse health systems and interventional trials targeting non–HDL-C to test risk reduction for incident metabolic diseases.
OBJECTIVE: To assess the prevalence of elevated non-HDL-C values in the general population in the absence of lipid-lowering drugs. Association with metabolic diseases and the risk to develop new ones were analyzed using data from electronic health records. PATIENTS AND METHODS: 820,318 subjects with at least two lipid measurements between 2012 and 2016, and without lipid-lowering drug treatment, were included. Once selected, non-HDL-C and LDL-C + 30 were calculated, with non-HDL-C being defined as elevated if it was higher than LDL-C + 30 mg (4.00 mmol) on two conse
3. Simplifying meal bolusing with Omnipod 5: real-world outcomes to support simplified user-initiated insulin dosing strategies in managing type 1 or type 2 diabetes.
Among 195,529 Omnipod 5 users across four countries, simplified meal bolusing achieved non-inferior glycemic outcomes to carbohydrate counting in T1D and T2D: modestly lower TIR offset by lower hypoglycemia (TBR) and small mean glucose differences. Findings support simplifying user-initiated boluses without compromising control.
Impact: This study provides practice-changing, real-world evidence that carbohydrate counting is not required to achieve strong glycemic outcomes with AID, lowering barriers to adoption and equity in diabetes technology.
Clinical Implications: Clinicians can endorse simplified meal bolusing with Omnipod 5, focusing education on safe use and settings optimization rather than precise carbohydrate counting; access to AID should not hinge on carb-counting proficiency.
Key Findings
- 195,529 users analyzed (92% T1D) across US, UK, France, and Germany
- In T1D, simplified bolusing vs carb counting: −2.5% TIR, −0.19% TBR, +5 mg/dL mean glucose; all endpoints met noninferiority
- In T2D, differences were −1.7% TIR, −0.05% TBR, +4 mg/dL mean glucose; noninferiority met
- Results were consistent across age groups and countries; adjusted quantile regression accounted for user/device factors
Methodological Strengths
- Exceptional sample size with multinational real-world data
- Adjusted analyses (quantile regression) incorporating user characteristics and device settings/usage
Limitations
- Retrospective observational design with potential residual confounding and selection bias
- Short observation window (3 months) limits assessment of long-term outcomes and sustainability
Future Directions: Prospective or randomized evaluations of simplified bolusing strategies over longer durations, assessing patient-reported outcomes, adherence, and hypoglycemia/hyperglycemia events.
AIMS: Automated insulin delivery (AID) systems typically require user‑initiated meal boluses. This study leveraged real-world data from Omnipod® 5 Automated Insulin Delivery System users to evaluate whether glycemic outcomes using simplified meal bolusing were non-inferior to carbohydrate counting. METHODS: A retrospective analysis of Omnipod 5 users in the US, UK, France, and Germany from 01/01/25 - 31/03/25 was conducted. Carbohydrate counting or simplified bolusing categorization was based on the proportion of boluses that used each individual's top five carbo