Abstract Objectives Long COVID is defined by persistent multi-organ symptoms without an alternative diagnosis. This study evaluated whether routine laboratory data collected under real-world conditions differentiate symptom-defined subgroups within an outpatient cohort and predict the course. Methods In this single-center cohort study of 564 Long COVID patients (1,452 visits), symptom phenotypes were derived by latent class analysis of 19 organ-system domains at baseline. Structure in 89 routine analytes was sought by k-means clustering, confirmed with alternative algorithms, and repeated after age residualization. Analytes were compared across phenotypes using Kruskal-Wallis tests with false-discovery-rate control and adjustment for confounders. Equivalence testing was performed against reference change values. The course was modeled by count models. Prediction of improvement used a nested cross-validated elastic net with a baseline-independent endpoint. Results Clustering showed no convincing structure (silhouette ≤0.09, two to eight clusters); the weak split tracked age (50.5 vs. 38 years) and vanished after residualization. Three age-independent phenotypes (n=503) had no laboratory correlate: none of 89 analytes survived FDR correction, with or without adjustment, or across 50 pseudo-class draws. Equivalence held for 86/89 analytes within ±0.5 SD and 63/70 within the reference change value. Symptom burden declined (median 4 to 2 organ systems, p=2.5 × 10 −19 ); however, no analyte predicted the course (AUC 0.46). Conclusions Routine laboratory variation in this cohort was dominated by age and decoupled from symptom phenotype in both directions. Routine testing in a real-world setting could not provide an age-independent diagnostic, subtyping, or prognostic signature in a cohort of patients suffering from Long COVID.
Abstract Objectives Clinical translation of extracellular vesicle (EV)-based diagnostics is constrained by preanalytical variability in separation workflows. We developed a preanalytical quality indicator framework for automated EV separation, anchored to IFCC, EFLM, and MISEV2023 standards. This proof-of-concept study establishes analytical prerequisites; validation in disease cohorts is required before clinical implementation. Methods Paired serum and plasma from 10 healthy controls were processed by EXODUS, Summit, and manual SEC. Endpoints included NTA yield, particle-to-protein ratio (PPR), intra-matrix reproducibility (CV%), and tetraspanin preservation (CD9, CD63, CD81) by SP-IRIS. Results EXODUS achieved the highest particle concentration and PPR, exceeding SEC-based methods by 5- to 10-fold. Summit showed the lowest process variation (CV% <20 %) in serum (18.9 %) and plasma (16.5 %), satisfying IFCC-EFLM robustness criteria. Manual SEC showed the highest variability (serum: 55.3 %; plasma: 45.5 %). SP-IRIS revealed comparable tetraspanin detection across EXODUS and Summit, with qEV showing lower intensity. Summit confirmed superior scalability (4–48 samples/run, 10–15 min hands-on time). Conclusions This proof-of-concept framework establishes analytical prerequisites. No single platform optimizes all parameters; selection should be driven by intended application: EXODUS for sensitivity-limited discovery, Summit for analytical standardization, manual SEC for morphological reference. Validation in disease cohorts is required before clinical implementation.
Abstract Objectives Spurious hypoxemia – a low PaO 2 with preserved, directly measured oxyhemoglobin saturation (sO 2 ) – is traditionally attributed to extreme leukocytosis or thrombocytosis (“leukocyte larceny”). We assessed how often it instead reflects PaO 2 temperature reporting or neonatal fetal-hemoglobin physiology. Methods We retrospectively analyzed 7,473 arterial blood-gas specimens from 2,647 patients (September 2020–January 2026) with co-oximetry-derived sO 2 , complete blood count, and metabolic panel data; regression analyses used the 7,431 specimens reported within 80 min of collection. Spurious hypoxemia was defined as PaO 2 <60 torr (8.0 kPa) referenced to 37 °C at measured pH, with sO 2 ≥92 %. Firth logistic regression, Kaplan–Meier analysis, and receiver-operating-characteristic curves evaluated temperature correction, neonatal status, and extreme leukocytosis (≥50 × 10 9 /L) or thrombocytosis (≥500 × 10 9 /L). Results Spurious hypoxemia affected 74 specimens (1.0 %). Among hypothermic specimens, the apparent rate fell from 3.5 % with PaO 2 reported at patient temperature to 0.9 % when referenced to 37 °C. Rates were higher in normothermic infants than adults (5.4 vs. 0.3 %), consistent with fetal hemoglobin. Extreme leukocytosis was associated with the outcome (adjusted odds ratio 20.2; 95 % confidence interval 5.6–60.9), but all four events occurred in neonates. Longer collection-to-reporting time was modestly associated (odds ratio 1.45 per 10 min), mainly in neonates. No non-neonatal specimen with a white-cell count ≥30 × 10 9 /L met the outcome (0/139), and extreme thrombocytosis showed no association (0/126). Conclusions Apparent spurious hypoxemia was more often explained by temperature reporting and fetal hemoglobin than oxygen metabolism by leukocytes; the leukocytosis signal could not be separated from neonatal physiology. Laboratories should identify the PaO 2 reporting temperature, and clinicians should consider fetal-hemoglobin physiology before invoking leukocyte-mediated hypoxemia in neonates.
Abstract Objectives Bile acids (BAs) play crucial roles in digestion and metabolic regulation, and their composition is increasingly recognised as relevant to health and disease. This study aimed to establish population-based concentration ranges (2.5th–97.5th percentiles) for 24 individual serum BAs and assess the influence of demographic and lifestyle factors in a large cohort of Swedish adults. Methods Serum samples from 931 Swedish adults were analysed using liquid chromatography–high-resolution mass spectrometry. Population-based concentration ranges were calculated for individual BAs, and associations with sex, age, BMI, sampling time, smoking, Swedish snuff use, serum cotinine, alcohol consumption, and ancestry were evaluated. Results We found extensive interindividual variability in BA concentrations. The most abundant BAs were glycine-conjugated forms, while primary unconjugated BAs exhibited the largest concentration ranges, differing by over 1,000-fold between individuals. Sex-specific patterns emerged, with men showing significantly higher levels of unconjugated BAs. Age and BMI were negatively associated with taurine-conjugated BAs, and weak time-of-day variations were observed for several conjugated BAs. Lifestyle factors such as smoking, assessed through self-report and serum cotinine, showed stronger associations with BA levels than alcohol or ancestry, notably with increased murideoxycholic acid (MDCA) and reduced conjugated BA levels among nicotine users. Conclusions These findings provide robust serum concentration ranges and highlight key variables that may confound BA biomarker interpretations, emphasising the need to consider sex, age, BMI, and lifestyle in future clinical and epidemiological applications.
Abstract Objectives To evaluate the analytical performance and clinical classification value of a novel sandwich testosterone immunoassay, particularly for biochemical testosterone elevation in women and deficiency in male hemodialysis patients, using LC–MS/MS as the reference method. Methods A total of 387 residual clinical serum samples were compared using Mindray TESTO, Siemens TSTII, and liquid chromatography–tandem mass spectrometry (LC–MS/MS). Among these, 226 had testosterone concentrations <1.0 ng/mL, and 133 reproductive-age women were evaluated for biochemical hyperandrogenism classification. An additional 119 hemodialysis samples (94 men, 25 women) were assessed, with the 94 men evaluated for biochemical testosterone deficiency classification. Analyses included Passing–Bablok regression, Bland–Altman analysis, and independent imprecision, linearity, and interference studies. Results The Mindray TESTO assay showed total CVs of 1.68–2.71 % and good linearity (r=0.9997). Cross-reactivity was generally low, although phenylbutazone produced biases exceeding ±10 %. Both immunoassays correlated strongly with LC–MS/MS, with Mindray TESTO showing higher correlation and narrower limits of agreement in low-concentration samples (r=0.9603 vs. 0.8920). In women of reproductive age, Mindray TESTO demonstrated higher sensitivity (88.89 vs. 11.11 %) and κ agreement (0.784 vs. 0.189) for biochemical hyperandrogenism classification. In male hemodialysis patients, it achieved higher accuracy (94.68 vs. 88.30 %) and κ agreement (0.838 vs. 0.680) for biochemical testosterone deficiency classification. The ROC AUCs did not differ significantly (0.966 vs. 0.956; p=0.2776). Conclusions The novel sandwich immunoassay showed improved agreement with LC–MS/MS at low testosterone concentrations and favorable performance in classifying biochemical testosterone elevation and deficiency.
