Gestational diabetes mellitus (GDM) is a prevalent condition requiring accurate patient education, yet the reliability and readability of large language models (LLMs) in this context remain uncertain. This study evaluated the performance of four LLMs—ChatGPT-4o, Gemini 2.5 Pro, Grok 3.0, and DeepSeek R-1—using 25 patient-oriented questions derived from clinical scenarios. Seven endocrinologists independently rated the responses with the modified DISCERN (mDISCERN) instrument and the Global Quality Score (GQS). Readability was analyzed using the Flesch Reading Ease (FRES), Flesch–Kincaid Grade Level (FKGL), Gunning Fog Index (GFI), Coleman–Liau Index (CLI), and Simple Measure of Gobbledygook (SMOG), while lexical diversity was assessed through type–token ratio (TTR). Grok and Gemini obtained the highest mDISCERN and GQS scores, whereas ChatGPT performed significantly lower (p < 0.05). DeepSeek generated the most readable outputs, while Grok provided the longest and most complex responses. All models scored below the FRES threshold of 60 recommended for lay audiences. Response length showed strong positive correlations with mDISCERN and GQS, while TTR was inversely related to quality but positively associated with readability. These findings highlight variability among LLMs in GDM education and emphasize the need for model-specific improvements to ensure reliable patient-facing health information.

Background and objective: Although many studies on the Diabetes Eating Problem Survey–Revised (DEPS-R) in adolescents with type 1 diabetes mellitus (T1D), the number of studies validating this questionnaire in adults with T1D is limited. Therefore, this study aimed to examine the factor structure of the Turkish version of the DEPS-R in adults with T1D and internal consistency and construct validity. Methods: A total of 100 patients with T1D, ages 18–50 years, completed the DEPS-R and EDE-Q. In addition to tests of validity, confirmatory factor analysis was conducted to investigate the factor structure of the 6-item Turkish version of DEPS-R. Results: The Cronbach’s alpha coefficient of the DEPS-R Turkish version was 0.77, suggesting good internal consistency. The median (IQ) DEPS-R score was 15.0 (13.0) among all participants. DEPS-R score was significantly correlated with BMI (r = 0.210; p < 0.05) and EDE-Q (r = 0.586; p < 0.01). There was no correlation between the HbA1c values of participants and neither EDE-Q nor DEPS-R scores. The confirmatory factor analysis results show that the three-factor model was a good fit. Conclusion: A short, self-administered diabetes-specific screening tool for disordered eating behavior is recommended be used routinely in the clinical care of adults with T1D, and Turkish version of DEPS-R has acceptable internal consistency and construct validity in adults with T1D. Level of evidence: Level V, descriptive study. Clinicaltrials.gov registration number NCT05346679/ 21.04.2022 (retrospectively registered)

Background/aim: To describe Graves’ Disease (GD) associated with COVID-19 infection (COVID) or its vaccines (VAC) and to compare the clinical presentations, laboratory parameters, and short-term clinical course of the disease among different etiology groups (COVID, VAC, and GD control). Materials and methods: Included in this multicenter matched case–control, retrospective cohort study were 239 patients with newly diagnosed (n = 196) or recurrent GD (n = 43) associated with COVID (n = 79) or VAC (n = 160). Each case was matched (1:1) with a control who had been diagnosed with GD prior to COVID. Results: The median age of the entire group was 42 years (female:male = 137:102). Both the COVID (4.6-fold) and VAC (4.1-fold) groups demonstrated higher TSH receptor antibody (TRAb) titers (p < 0.001) compared with the control group (3.5-fold), as well as a higher proportion of recurrent cases. At baseline, the COVID group had higher free triiodothyronine (fT3) levels than the other groups. Graves orbitopathy (GO) was observed in 60 patients (12.6%), with a higher frequency in classical GD (18.4%). At baseline, the variables associated with thyrotoxicosis severity (defined as fT3 levels) were younger age, higher thyroid gland volume (TGV), and etiology, with the COVID and, to a lesser extent, VAC groups presenting with higher fT3 levels. The variables associated with GO were higher TGV, TRAb titers, and smoking, while no association with etiology was identified. Conclusion: The clinical course was similar in all groups other than in some laboratory findings. Although the frequency of GO associated with COVID and VAC was lower, the proportion of cases with a Clinical Activity Score of ≥3 was higher compared to GD. This pattern suggests a potentially stronger immunologic trigger in these cases.

Rationale: Insulin antibody-mediated insulin resistance is a rare autoimmune mechanism that can cause severe hyperglycemia. Patient concerns: A 52-year-old male patient was admitted to our hospital with complaints of polydipsia, polyuria, and weight loss (8 kg in 6 months). He was diagnosed with type 2 diabetes mellitus at age 33. Diagnoses: Type 2 diabetes mellitus with antibody-mediated insulin resistance syndrome. Interventions: Rituximab infusion in a 500 mg dose was given 2 times in a 2-week interval. Outcomes: A significant response was achieved 1 month later with fasting plasma glucose: 120 mg/dL (6.7 mmol/L), glycosylated hemoglobin A1c: 7.6% (59.6 mmol/mol), and anti-insulin antibodies: 0.001 U/mL. Clinically improved response persisted for about 6 months. Lessons: Clinicians should be aware of the antibody-mediated insulin resistance, recognize suggestive signs and symptoms, pursue appropriate diagnostic evaluation, and treatment approach.

Background: Minimal hepatic encephalopathy can only be detected by specific psychometric or neuropsychological tests. We aimed to determine the prevalence of minimal hepatic encephalopathy in a hepatology outpatient clinic of a tertiary center. Methods: A total of 82 patients with chronic liver disease were involved prospectively in this study. Control groups consisted of healthy volunteers (n = 123) and chronic renal failure patients (n = 28). We used 2 different methods to detect minimal hepatic encephalopathy. First method was a battery of 5 psychometric tests (number connection tests A and B, digit symbol test, serial dot test, line tracing test) which was filled by all patients. The second method was critical flicker frequency test. Both methods were used in the whole group (n = 233). We applied linear regression analysis to the results of psychometric tests of healthy volunteers to establish equations to calculate the expected values of each test. Test results of the patients were evaluated according to the expected results obtained from these equations. Results: The prevalence of minimal hepatic encephalopathy detected by psychometric tests and critical flicker frequency test was 13% and 14%, respectively. When the positivity of both tests was deemed necessary to diagnose minimal hepatic encephalopathy, the rate of minimal hepatic encephalopathy was 3.6% (n = 3) in a chronic liver disease patient group. Conclusion: Minimal hepatic encephalopathy is a difficult clinical condition to diagnose, and it is more appropriate to use psychometric tests and critical flicker frequency test together.

Background and Aims: The age‐at‐onset is of great importance in the heterogeneity of Type 1 diabetes mellitus (T1DM). This study was designed to define clinical and laboratory differences between child–adolescent‐onset and adult‐onset T1DM at presentation and during follow‐up and determine the predicting factors for developing microvascular and macrovascular complications. Material and Methods: This retrospective observational study evaluated T1DM patients who were followed in the diabetes outpatient clinic between January 1, 2000, and December 31, 2019. Results: The study cohort included 490 individuals with T1DM (54.3% female, 58.8% adult‐onset, and median follow‐up: 5 years). In the adult‐onset group, baseline C‐peptide and GADA prevalence were higher, whereas presentation with ketoacidosis was 2.3‐fold lower compared to the child–adolescent‐onset group ( p < 0.001). During follow‐up, the adult‐onset group had a 2.4‐fold higher overweight/obesity ( p < 0.001) and 1.7‐fold higher dyslipidemia/hyperlipidemia ( p = 0.002) than the child–adolescent‐onset group. In multivariate analysis, fasting glucose ( p = 0.024) in adult‐onset, dyslipidemia/hyperlipidemia ( p = 0.037) in child–adolescent‐onset, and diabetes duration ( p = 0.008 and p = 0.007) and hypertension ( p = 0.001 and p = 0.01) in both groups were associated with increased risk of microvascular complications, whereas age‐at‐onset ( p = 0.024), dyslipidemia/hyperlipidemia ( p = 0.03), nephropathy ( p = 0.003), and neuropathy ( p = 0.001) in adult‐onset and age ( p = 0.002) and triglycerides ( p = 0.013) in child–adolescent‐onset groups were associated with increased risk of macrovascular complications. The cutoff C‐peptide levels at baseline predicted microvascular complications in the whole cohort and adult‐onset group were defined as 0.383 ng/mL ( p < 0.001) and 0.41 ng/mL ( p = 0.001), respectively. In the Kaplan–Meier analysis, C‐peptide (< 0.383 ng/mL) but not age‐at‐onset predicted future development of microvascular and macrovascular complications ( p = 0.003 and p = 0.032). Conclusion: Clinical presentation and prognosis differ in adult‐onset and child–adolescent‐onset T1DM. Low initial C‐peptide may predict the development of microvascular and macrovascular complications.