Jaak Billen

Equilibrium dialysis followed by liquid chromatography–tandem mass spectrometry is considered the gold standard for the isolation and quantification of free hormones. However, this technique has not been described for the analysis of free 25-hydroxyvitamin D3. This study investigated analytical challenges, focusing on stability and adsorption of 25-hydroxyvitamin D3 in aqueous buffer environments, to be considered prior to method development of equilibrium dialysis for free vitamin D. Experiments revealed that 25-hydroxyvitamin D3 is highly prone to adsorption, and to a minor extent degradation, resulting in severe analyte loss already upon introduction of 25-hydroxyvitamin D3 into the aqueous environment, far exceeding observations in serum where protein binding provides significant protection. While chemical degradation only slightly contributes, strong surface adsorption accounts for the majority of 25-hydroxyvitamin D3 loss observed. In contrast, the addition of methanol to the buffer environment completely mitigated analyte loss, confirming the severe adsorption of 25-hydroxyvitamin D3 to recipient surfaces. However, incorporating methanol as a buffer component fundamentally altered experimental conditions, thereby potentially disrupting the physiological equilibrium between 25-hydroxyvitamin D3, vitamin D-binding protein, and albumin, and by extension the free fraction of vitamin D. Consequently, the large extent of, primarily, adsorption of 25-hydroxyvitamin D3 in aqueous environments represents an important consideration in setting up equilibrium dialysis experiments for free vitamin D. These findings additionally highlight a critical need for novel analytical strategies that prevent adsorption while maintaining physiological equilibrium in the assessment of free vitamin D.

Background: Urinalysis is a standard clinical test that includes the microscopic examination of urinary sediment to identify formed elements. Manual evaluation by laboratory technicians is time-intensive and subject to human error. Automated analysis using digital microscopy images presents a potential alternative. This study evaluates the integration of a deep learning approach to automatically classify urinary sediment images in the clinical laboratory, including independent prospective validation of its performance. Methods: An annotated data set comprising 13 classes of urinary sediment elements was created from a database of Sysmex UD-10 digital microscope images. An EfficientNet-based model was trained and tested across three experimental scenarios to evaluate the effects of data collection strategies on performance. Uncertainty calibration was examined. The model’s robustness and interpretability were examined using gradient-weighted class activation mapping (Grad-CAM) to visualize influential image regions and t-distributed stochastic neighbor embedding (t-SNE) to analyze learned feature embeddings. Lastly, a graphical user interface was developed for a prospective evaluation in the laboratory. Results: The model achieved approximately 97% overall accuracy on the test set. Experiments revealed sensitivity to data set variability, suggesting that performance may improve by integrating additional training examples. Confidence scores aligned with accuracy, and interpretability analyses showed that the model focused on relevant image regions and learned embeddings demonstrated clear class separation. In the prospective evaluation, top 1 and top 3 accuracies decreased to approximately 78% and 92%, respectively. Conclusions: Our results indicate that a lightweight deep learning model can achieve high performance in classifying urine particles. Analysis of discrepancies between retrospective and prospective evaluations provides important insights toward reliable clinical application.

Background/Objectives: To evaluate diabetes risk perception in women with prior gestational diabetes mellitus (GDM) and prediabetes in early postpartum. Methods: Secondary analysis of a multi-center randomized controlled trial assessing the effectiveness of a mobile-based postpartum lifestyle intervention in women with prediabetes after GDM. Data were collected from the Risk Perception Survey for Developing Diabetes at baseline (6–16 weeks postpartum) and one year post-randomization. Logistic regression was used to analyze the difference between the intervention and control groups on diabetes risk estimation. Results: Among 165 women with prediabetes in early postpartum (mean age: 32.1 years, mean BMI: 27.3 kg/m²), 58.9% (96) adequately estimated their diabetes risk (moderate or high chance) at baseline. These women smoked less often [2.06% (2) vs. 10.3% (7), p = 0.034], reported less anxiety (11.6 ± 3.0 vs. 12.6 ± 3.5, p = 0.040), and reported fewer symptoms of depression [30.9% (21) vs. 15.6% (15), p = 0.023] compared to women who underestimated their risk. At one year, 58.3% (95) of all women adequately estimated their diabetes risk. In the intervention group, 50.6% (41) adequately estimated their risk at baseline, increasing to 56.8% (46) by the end of the intervention after one year (p = 0.638). In the control group, a higher proportion of women adequately estimated their risk at baseline [67.1% (55), (p = 0.039)], which decreased to 59.8% (49) at one year (p = 0.376), with no significant difference in risk perception between the groups at one year (p = 0.638). Conclusions: Almost 60% of this high-risk population adequately estimated their diabetes risk, with no significant impact of the lifestyle intervention on risk perception.

Aims: To examine the association between depressive symptoms and metabolic profile in women with prior gestational diabetes mellitus (GDM) and early postpartum prediabetes, and to explore whether a mobile-based lifestyle intervention affected mental health outcomes. Methods: Secondary, exploratory analysis of a multi-centric randomized controlled trial (MELINDA), evaluating a mobile-based lifestyle intervention versus standard follow-up (control group) in women with prediabetes after GDM. The analysis included 166 participants who completed the Center for Epidemiologic Studies–Depression (CES-D) questionnaire [score of ≥16 being suggestive for (sub)clinical depression] at baseline (6–16 weeks postpartum) and one year post-randomization. Results: At one year, 26.5% of women (n = 44) reported depressive symptoms, with no significant difference between the intervention and control groups (30.5% vs. 22.6%, p = 0.293). Women with depressive symptoms (symptomatic women) were younger (30.9 ± 4.9 vs. 32.5 ± 3.8 years, p = 0.033) and were less often highly educated (61.4% vs. 80.3%, p = 0.028). At baseline, symptomatic women had a higher rate of metabolic syndrome (38.6% vs. 21.9%, p = 0.044), higher LDL-cholesterol [3.2 ± 0.8 vs. 2.8 ± 0.8 mmol/L, p = 0.009], lower quality of life (lower SF-36 scores, p < 0.050) and a higher level of anxiety based on the STAI-6 questionnaire (14.5 ± 3.6 vs. 11.2 ± 2.6, p < 0.001). These differences persisted at one year postpartum with worse metabolic profile, more anxiety and lower quality of life in symptomatic women. Conclusions: Depressive symptoms are common in women with prediabetes in early postpartum after GDM and are associated with a persistent worse metabolic profile, increased anxiety and lower quality of life postpartum. The mobile-based lifestyle intervention did not improve mental health.

Objectives: To compare clinical laboratory workflows for the assessment of androgens in men, focusing on total testosterone (T), sex hormone-binding globulin (SHBG) and free T, in clinical laboratories throughout Europe. Methods: An internet-based survey that included questions related to pre-analytical, analytical and post-analytical phases of androgen measurements was distributed between December 2022 and December 2023 by clinical laboratory/chemistry and endocrine societies. A total of 124 unique records from clinical laboratories in 27 European countries were analyzed. Results: Pre-analytical requirements for total T are subject to improvement as less than half of clinical laboratories recommended adequate morning sampling time and/or sampling in a fasting state. Total T was predominantly quantified using enzyme-linked immunoassay (IA) on automated platforms, with only one in four centers using mass spectrometry (MS), while SHBG was exclusively measured by IA. Additionally, free T was used by a majority of clinical laboratories, mainly reported as approximation by calculation of free T (cFT) using the Vermeulen formula. Generally, age-stratification was the preferred means of reporting reference ranges for total T, SHBG and cFT. However, considerate variability was observed in reported lower and upper limits, leading to the necessity of interpreting test results against assay-specific reference ranges, thereby hindering comparability of results between clinical laboratories. Conclusions: Our survey highlights significant inter-laboratory variability for the assessment of androgen status in men, implying non-commutability of clinical test results between different centers. In addition, we observed poor adherence to pre-analytical recommendations. These findings advocate for continued harmonization efforts of measurement procedures for SHBG and total/free T.

The hormone hepcidin plays an important role in intestinal iron absorption and cellular release. Cord blood hepcidin values reflect fetal hepcidin status, at least at the time of delivery, but are not available for the Belgian population. Therefore, we aimed (1) to provide the first data on cord blood hepcidin levels in a Belgian cohort and (2) to determine variables associated with cord blood hepcidin concentrations. A cross-sectional, observational study was performed at the University Hospital Leuven, Belgium. Cord blood samples were analyzed using a combination of weak cation exchange chromatography and time-of-flight mass spectrometry. Descriptive statistics, Spearman correlation tests, and Mann–Whitney U tests were performed. In total, 61 nonhemolyzed cord blood samples were analyzed. The median hepcidin level was 17.6 μg/L (IQR: 18.1; min-max: 3.9–54.7). A moderate correlation was observed between cord blood hepcidin and cord blood ferritin (r = 0.493) and hemoglobin (r = −0.342). Cord blood hepcidin was also associated with mode of delivery (p = 0.01), with higher hepcidin levels for vaginal deliveries. Nonetheless, larger studies are needed to provide more evidence on the actual clinical value and benefit of cord blood hepcidin measurements.

Background and Objectives Rebaudioside A, a steviol glycoside, is deglycosylated by intestinal microflora prior to the absorption of steviol and conjugation to steviol glucuronide. While glucose-lowering properties are observed for rebaudioside A in mice, they have been attributed to the metabolites steviol and steviol glucuronide. We aimed to characterize the pharmacokinetic and pharmacodynamic properties of rebaudioside A and its metabolites in patients with early-onset type 2 diabetes mellitus (T2DM). Methods This randomized, placebo-controlled, open-label, two-way crossover trial was performed in subjects with T2DM on metformin or no therapy at the University Hospitals Leuven, Belgium. Following oral rebaudioside A (3 g), plasma concentrations of rebaudioside A, steviol and steviol glucuronide were determined. The effect on glucose homeostasis was examined by an oral glucose tolerance test (OGTT) performed 19 h following rebaudioside A administration, i.e. the presumed time of maximal steviol and steviol glucuronide concentrations. The primary pharmacodynamic endpoint was the difference in area under the blood glucose concentration–time curve during the first 2 h of the OGTT (AUC_{Glucose(0–2h)}) for rebaudioside A vs. placebo. Results In total, 30 subjects [63.5 (57.8–69.0) years of age, 86.7% male] completed the trial. Rebaudioside A was detected as early as 1 h after administration in nearly all subjects. As expected, steviol and steviol glucuronide reached their maximal concentrations at 19.5 h following rebaudioside A administration. Rebaudioside A did not lower the AUC_{Glucose(0–2h)} compared to placebo (− 0.7 (95% CI − 22.3; 20.9) h·mg/dL, P = 0.95). Insulin and C-peptide concentrations were also comparable between both conditions (P > 0.05). Conclusion Rebaudioside A is readily absorbed after oral administration and metabolized to steviol and steviol glucuronide. However, no effect on glucose nor insulin or C-peptide excursion was observed during the OGTT at the time of maximal metabolite concentrations. Thus, no antidiabetic properties of rebaudioside A could be observed in patients with T2DM after single oral use. Clinical Trial Registration Registered on ClinicalTrials.gov (NCT03510624).

Nutrition is important during pregnancy for offspring health. Gestational vitamin D intake may prevent several adverse outcomes and might have an influence on offspring telomere length (TL). In this study, we want to assess the association between maternal vitamin D intake during pregnancy and newborn TL, as reflected by cord blood TL. We studied mother–child pairs enrolled in the Maternal Nutrition and Offspring’s Epigenome (MANOE) cohort, Leuven, Belgium. To calculate the dietary vitamin D intake, 108 women were asked to keep track of their diet using the seven-day estimated diet record (EDR) method. TL was assessed in 108 cord blood using a quantitative real-time PCR method. In each trimester of pregnancy, maternal serum 25-hydroxyvitamin D (25-OHD) concentration was measured. We observed a positive association (β = 0.009, p-value = 0.036) between newborn average relative TL and maternal vitamin D intake (diet + supplement) during the first trimester. In contrast, we found no association between average relative TL of the newborn and mean maternal serum 25-OHD concentrations during pregnancy. To conclude, vitamin D intake (diet + supplements), specifically during the first trimester of pregnancy, is an important factor associated with TL at birth.

The aims of the ‘Mobile-based lifestyle intervention in women with glucose intolerance after gestational diabetes mellitus (GDM)’ study (MELINDA) are: (1) to evaluate the prevalence and risk factors of glucose intolerance after a recent history of GDM; and (2) to evaluate the efficacy and feasibility of a telephone- and mobile-based lifestyle intervention in women with glucose intolerance after GDM. This is a Belgian multicenter randomized controlled trial (RCT) in seven hospitals with the aim of recruiting 236 women. Women in the intervention group will receive a blended program, based on one face-to-face education session and further follow-up through a mobile application and monthly telephone advice. Women in the control group will receive follow-up as in normal routine with referral to primary care. Participants will receive an oral glucose tolerance test (OGTT) one year after baseline. Primary endpoint is the frequency of weight goal achievement (≥5% weight loss if pre-pregnancy BMI ≥ 25 Kg/m² or return to pre-gravid weight if BMI < 25 Kg/m²). At each visit blood samples are collected, anthropometric measurements are obtained, and self-administered questionnaires are completed. Recruitment began in May 2019.