Schilddrüsenerkrankungen sind in Europa von hoher Relevanz, und obwohl Jodmangel nach wie vor die am besten bekannte ernährungsbedingte Ursache ist, gibt es Evidenz, dass andere Mikronährstoffe – darunter Selen, Eisen, Zink und Vitamin D – die Schilddrüsenfunktion beeinflussen. Sowohl ein Mangel als auch ein Überschuss können die Hormonsynthese, die Immunregulation und damit die Therapieergebnisse beeinträchtigen. Ernährungsgewohnheiten, insbesondere die Nichtaufnahme der relevanten Lebensmittel, das Abhandensein der angezeigten Nährstoffe in Lebensmitteln sowie der hohe Verzehr von stark verarbeiteten Lebensmitteln, können zu einer suboptimalen Mikronährstoffzufuhr beitragen. Jod ist für die Produktion von Schilddrüsenhormonen unerlässlich. Selen unterstützt die antioxidative Abwehr und den Hormonstoffwechsel, Eisen ist für die Aktivität der Schilddrüsenperoxidase erforderlich, Zink trägt zur Hormonregulation bei und Vitamin D könnte eine Rolle bei Autoimmunerkrankungen der Schilddrüse spielen, obwohl die Kausalität noch unklar ist. Alle haben gemeinsam: sowohl eine zu geringe als auch eine zu hohe Zufuhr birgt gesundheitliche Risiken für die Schilddrüse. Sogenannte Antinährstoffe wie Goitrogene sind bei ausreichender Jodzufuhr im Allgemeinen unbedenklich, teilweise sogar gesundheitsförderlich, und das Wissen um die Handhabung reduziert mögliche adverse Auswirkungen. Darüber hinaus sind Wechselwirkungen zwischen Nährstoffen – insbesondere mit komplexen Lebensmitteln und Schilddrüsenarzneien wie Levothyroxin – klinisch relevant hinsichtlich Bioverfügbarkeit und erfordern Aufmerksamkeit bezüglich des Aufnahmezeitpunkts und der Zusammensetzung. Insgesamt ist die Aufrechterhaltung eines ausgeglichenen Mikronährstoffstatus hoch relevant für die Gesundheit der Schilddrüse, jedoch wird eine routinemäßige Supplementierung über bestätigte Mangelzustände hinaus nicht durch Evidenz gestützt. Regelmäßige Routinekontrollen können helfen, einen Mangel frühzeitig zu erkennen und zu behandeln und bestimmten Schilddrüsenerkrankungen vorzubeugen.
Abstract Background Night shift workers are more likely to exhibit unfavorable dietary behaviors than day workers. However, these differences remain insufficiently explored within the European workforce. This study aims to examine differences in dietary behaviors (food intake frequency, eating frequency) by night shift exposure (history, frequency, duration) to investigate eating frequency, meal timing, and food choice determinants during night shifts, and to explore gender differences across Europe. Methods Data were collected via an online survey (May 2024-January 2025) in eight countries (Austria, Germany, Denmark, Greece, Italy, the Netherlands, Poland, Spain). Participants self-reported sociodemographics, occupational sector, current work schedules, shift work history, lifestyle characteristics, and dietary behaviors. A shortened Food Frequency Questionnaire (FFQ) assessed dietary intake, alongside questions on eating rate and frequency on work and non-work-days. Current night shift workers additionally reported eating frequency, timing, and food choice determinants during night shifts. Analyses compared dietary behaviors by night shift exposure (current, former, vs. day worker). Among current night shift workers, associations with night shift frequency (nights/month) and duration (years) were examined. Secondary analyses were stratified by gender. Results A total of 6,260 individuals were included (mean age 40.7, SD 10.7; 50.5% female). Overall, 60.4% were current night shift workers, 19.6% former night shift workers, and 20% day workers. Compared to day workers, current night shift workers reported significantly faster eating rates (OR = 1.21, 95% CI: 1.06–1.37), more frequent intake of sugar-sweetened (OR = 1.30, 95% CI: 1.14–1.48) and caffeinated beverages (OR = 1.14, 95% CI: 1.01–1.30), and lower fruit intake (OR = 0.87, 95% CI: 0.77–0.98). Higher monthly night shift load and longer duration of night work were associated with less favorable dietary patterns. Most current night shift workers reported one to two eating occasions per night, typically at the beginning or middle of their shift. Food choices were primarily driven by appetite, time, and food availability. Gender differences were observed only in food choice determinants. Conclusions Night shift and day workers in Europe showed differences in dietary behaviors, particularly in sugar-sweetened beverage intake and eating rate. These findings highlight the need for targeted interventions to promote healthy eating among shift working populations.
Abstract The study focuses on the integration of Human Resource Information Systems (HRIS) into three types of software solutions (ERP, ATS, LMS) and their impact on human resource management. A qualitative comparative approach was employed, involving semistructure d interviews with 10 HR, IT, and line managers in Poland and Austria. The review yielded four central themes: efficiency and effectiveness, user experiences, integration challenges, and recommendations. Integration was cited as a way to improve data quality, expedite decision making, and enable advanced workforce analytics - especially in digital-savvy organisations. Resistance to users shifting into consensus due to the use of isomorphic representations was recorded in the Technology Acceptance Model. The small sample size and regional nature of the study constitute its limitations. It helps managers understand how integrated HR systems may affect organisational efficiency and coordination. This study combines several theoretical perspectives to reveal the nature and strategies of HRIS integration.
The increasing availability of sensor data enables data-driven monitoring and anomaly detection in distributed and cyber-physical systems. In unsupervised time series anomaly detection, reconstruction-based deep learning models are commonly employed, yet they implicitly assume that the training data is free of anomalies. Especially in large datasets, it can be difficult to ensure that the training data is indeed completely free of anomalies. This raises the question of how contaminated training data affects the anomaly detection performance quantified by the point-wise F1-score. To this end, we analyze the impact of increasingly contaminated training data sets on the F1-score using TimesNet, USAD and TranAD as state-of-the art models and a single Transformer block as a simple baseline model. The results show a striking sensitivity of several state-of-the-art unsupervised anomaly detectors to even small amounts of contamination in the training data. Models such as TranAD, Transformer, and TimesNet perform very well only when the training data set is completely clean. Once the training data contains 2.5–7.5% anomalous samples, their F1-scores drop by roughly 50%, revealing a strong dependence on uncontaminated training conditions. In contrast, USAD demonstrates substantially higher robustness to contamination. Even when 40% of the training samples are contaminated, USAD’s best F1-score remains above half of its clean-data performance. This indicates that USAD can tolerate heavy anomaly presence in the training data set without collapsing.
