Daily Endocrinology Research Analysis
Three high-impact endocrinology studies stood out: a national real-world evaluation showed multiple AI systems can safely triage diabetic retinopathy grading at scale; a secondary analysis of the STEP TEENS RCT found semaglutide 2.4 mg improves insulin sensitivity and cardiometabolic risk in adolescents with obesity; and a randomized crossover trial showed short-course low-dose hydrochlorothiazide under salt restriction raises serum calcium in nonsurgical hypoparathyroidism.
Summary
Three high-impact endocrinology studies stood out: a national real-world evaluation showed multiple AI systems can safely triage diabetic retinopathy grading at scale; a secondary analysis of the STEP TEENS RCT found semaglutide 2.4 mg improves insulin sensitivity and cardiometabolic risk in adolescents with obesity; and a randomized crossover trial showed short-course low-dose hydrochlorothiazide under salt restriction raises serum calcium in nonsurgical hypoparathyroidism.
Research Themes
- AI-enabled triage in diabetic retinopathy screening
- GLP-1 receptor agonist effects on insulin sensitivity in adolescent obesity
- Thiazide diuretics under salt restriction for hypoparathyroidism management
Selected Articles
1. Automated retinal image analysis systems to triage for grading of diabetic retinopathy: a large-scale, open-label, national screening programme in England.
In a national screening program with 202,886 encounters and ~1.2 million images, multiple CE-marked AI systems achieved high sensitivity for referable diabetic retinopathy (83.7–98.7%) and maintained equitable performance across age, sex, ethnicity, and deprivation subgroups. Sensitivity for moderate-to-severe NPDR and PDR was ≥95.8%, while performance for mild-to-moderate NPDR with referable maculopathy varied more; false-positive rates for no retinopathy ranged widely between vendors.
Impact: This is one of the largest real-world evaluations showing that AI can safely triage diabetic retinopathy grading at scale with equitable performance across diverse populations.
Clinical Implications: Health systems can deploy AI to triage for human grading, expand capacity, and accelerate detection of referable disease while monitoring vendor-specific false-positive behavior and maculopathy performance.
Key Findings
- Evaluated 8 CE-marked ARIAS on ~1.2 million images from 202,886 encounters with human grading reference.
- Sensitivity for referable diabetic retinopathy ranged 83.7–98.7%; for moderate-to-severe NPDR 96.7–99.8% and for PDR 95.8–99.5%.
- Performance was largely consistent across age, sex, ethnicity, and deprivation subgroups; maculopathy-related sensitivity showed greater variability.
- False-positive rates for no retinopathy varied widely between vendors (4.3–61.4%).
Methodological Strengths
- Large-scale real-world dataset with rigorous human reference grading.
- Head-to-head evaluation of multiple CE-marked algorithms with subgroup equity analyses.
Limitations
- Observational design without randomized workflow comparison or outcome follow-up.
- Vendor-specific variability in false positives and maculopathy sensitivity may affect downstream workload.
Future Directions: Prospective implementation studies comparing AI-first triage workflows vs standard care with cost-effectiveness, safety monitoring, and calibration for maculopathy detection.
BACKGROUND: The global prevalence of diabetes is rising, alongside costs and workload associated with screening for diabetic eye disease (diabetic retinopathy). Automated retinal image analysis systems (ARIAS) could replace primary human grading of images for diabetic retinopathy. We evaluated multiple ARIAS in a real-life screening programme. METHODS: Eight of 25 invited and potentially eligible CE-marked systems for diabetic retinopathy detection from retinal images agreed to participate. From 202 886 screening encounters at the North East London Diabetic Eye Screening Programme (between Jan 1, 2021, and Dec 31, 2022) we curated a database of 1·2 million images and sociodemographic and grading data. Images were manually graded by up to three graders according to a standard national protocol. ARIAS performance overall and by subgroups of age, sex, ethnicity, and index of multiple deprivation (IMD) were assessed against the reference standard, defined as the final human grade in the worst eye for referable diabetic retinopathy (primary outcome). Vendor algorithms did not have access to human grading data. FINDINGS: Sensitivity across vendors ranged from 83·7% to 98·7% for referable diabetic retinopathy, from 96·7% to 99·8% for moderate-to-severe non-proliferative diabetic retinopathy, and from 95·8% to 99·5% for proliferative diabetic retinopathy.
2. Effect of Semaglutide on Insulin Sensitivity and Cardiometabolic Risk Factors in Adolescents With Obesity: The STEP TEENS Study.
Among 193 adolescents without diabetes, semaglutide 2.4 mg led to greater reductions than placebo in fasting insulin (-33.6% vs -10.1%), HOMA-IR (-35.0% vs -5.3%), HbA1c, fasting glucose, triglycerides, LDL-C, total cholesterol, ALT, and waist-to-height ratio over 68 weeks. Improvements were larger in participants achieving ≥20% BMI reduction.
Impact: Extends adolescent obesity treatment evidence by demonstrating improvements in insulin sensitivity and multiple cardiometabolic risk factors beyond BMI reduction in a randomized setting.
Clinical Implications: Semaglutide 2.4 mg can be considered to improve insulin resistance and broader cardiometabolic risk in adolescents with obesity, with the greatest benefits accompanying substantial BMI reduction; monitoring of glycemic and hepatic parameters is warranted.
Key Findings
- Semaglutide reduced fasting insulin by 33.6% vs 10.1% with placebo (P=0.0012) and HOMA-IR by 35.0% vs 5.3% (P=0.0002).
- Improved HbA1c (P<0.0001), fasting plasma glucose (P=0.0181), triglycerides (P<0.0001), LDL-C (P=0.0105), total cholesterol (P<0.0001), ALT (-17.9% vs -3.3%; P=0.0232), and waist-to-height ratio (P<0.0001).
- Greater improvements were observed among participants achieving ≥20% BMI reduction.
Methodological Strengths
- Secondary analysis within a randomized, placebo-controlled phase 3a trial.
- Comprehensive cardiometabolic panel with consistent effects across measures.
Limitations
- Secondary analysis; insulin sensitivity assessed by surrogate (HOMA-IR) rather than clamp.
- Excluded youths with type 2 diabetes; generalizability may be limited.
Future Directions: Mechanistic studies with clamp-based assessments; longer-term outcomes on incident diabetes, hepatic steatosis, and cardiovascular risk in adolescents.
OBJECTIVE: This secondary analysis of the Semaglutide Treatment Effect in People with obesity (STEP) TEENS (NCT04102189) study investigated the effect of semaglutide 2.4 mg versus placebo on insulin sensitivity and cardiometabolic risk factors. RESEARCH DESIGN AND METHODS: The STEP TEENS phase 3a randomized study in adolescents (aged 12 to <18 years) with obesity demonstrated that once-weekly subcutaneous semaglutide 2.4 mg provided a significantly greater percentage reduction in BMI than placebo at week 68 (estimated difference -16.7 percentage points; P = 0.0001). This analysis investigated changes in insulin sensitivity and cardiometabolic risk factors from baseline to week 68. RESULTS: Overall, 193 participants without type 2 diabetes were included in the analysis. Participants receiving semaglutide 2.4 mg (n = 129) compared with those receiving placebo (n = 64) had greater reductions from baseline in fasting serum insulin (-33.6% vs. -10.1%; P = 0.0012), homeostatic model assessment for insulin resistance (HOMA-IR) score (-35.0% vs. -5.3%; P = 0.0002), glycemic measures (glycated hemoglobin:
3. Hydrochlorothiazide with salt restriction in nonsurgical hypoparathyroidism: a placebo-controlled single-blinded randomized crossover trial assessing efficacy and safety.
In adults with nonsurgical hypoparathyroidism stabilized on calcium/calcitriol and low-sodium diet, low-dose hydrochlorothiazide (12.5–25 mg/day) significantly increased serum calcium by day 5 and day 8 compared to baseline in a randomized, placebo-controlled crossover design. Secondary endpoints included urinary calcium excretion and sodium, with an optional 4-week open-label extension.
Impact: Provides randomized crossover evidence supporting a practical, low-dose thiazide plus salt restriction strategy to improve calcemia in nonsurgical hypoparathyroidism.
Clinical Implications: For patients with nonsurgical hypoparathyroidism on calcium/calcitriol, adding low-dose hydrochlorothiazide under salt restriction can raise serum calcium over a short period; clinicians should monitor electrolytes and urinary calcium and individualize dosing.
Key Findings
- Single-blinded randomized placebo-controlled crossover trial (n=26) under a controlled low-sodium diet.
- Low-dose hydrochlorothiazide (12.5–25 mg/day) significantly increased serum calcium at day 5 and day 8 compared with baseline (from 8.61±0.32 to 9.01±0.46 and 9.04±0.52 mg/dL).
- Secondary endpoints included 24-hour urinary calcium, fractional calcium excretion, urinary sodium, and bone turnover markers; an optional 4-week open-label HCTZ extension was offered.
Methodological Strengths
- Randomized placebo-controlled crossover design enabling within-subject comparisons.
- Dietary sodium controlled to standardize thiazide responsiveness.
Limitations
- Small single-center study with short intervention periods (7 days) and single-blind masking.
- Abstract does not report detailed secondary outcome magnitudes; longer-term efficacy and safety remain to be defined.
Future Directions: Multi-center, longer-duration randomized trials to quantify effects on urinary calcium, nephrocalcinosis risk, quality of life, and to refine dosing strategies.
In this single-center, randomized, placebo-controlled, single-blinded, crossover trial (Clinical Trial Registry-India: CTRI/2024/08/090041), we evaluated the efficacy and safety of low-dose hydrochlorothiazide (HCTZ) in adults with nonsurgical hypoparathyroidism (excluding autoimmune and obvious syndromic forms). After stabilization on fixed doses of calcium carbonate and calcitriol under a controlled low-sodium diet, 26 participants received HCTZ (12.5-25 mg/d) or placebo for 7 d, with a 15-d washout before crossover, followed by an optional 4-wk open-label HCTZ extension. The primary outcome was change in serum calcium; secondary outcomes included change in 24-h urinary calcium (UCa), fractional calcium excretion, urinary sodium, and bone turnover markers (β-CTX and P1NP). Compared to baseline (8.61 ± 0.32 mg/dL), serum calcium increased significantly with HCTZ at day 5 (9.01 ± 0.46 mg/dL) and day 8 (9.04 ± 0.52 mg/dL;