Machine Learning Prediction of Prediabetes in a Young Male Chinese Cohort with 5.8-Year Follow-Up.

The identification of risk factors for future prediabetes in young men remains largely unexamined. This study enrolled 6247 young ethnic Chinese men with normal fasting plasma glucose at the baseline (FPGbase), and used machine learning (Mach-L) methods to predict prediabetes after 5.8 years. The study seeks to achieve the following: 1. Evaluate whether Mach-L outperformed traditional multiple linear regression (MLR). 2. Identify the most important risk factors. The baseline data included demographic, biochemistry, and lifestyle information. Two models were built, where Model 1 included all variables and Model 2 excluded FPGbase, since it had the most profound effect on prediction. Random forest, stochastic gradient boosting, eXtreme gradient boosting, and elastic net were used, and the model performance was compared using different error metrics. All the Mach-L errors were smaller than those for MLR, thus Mach-L provided the most accurate results. In descending order of importance, the key factors for Model 1 were FPGbase, body fat (BF), creatinine (Cr), thyroid stimulating hormone (TSH), WBC, and age, while those for Model 2 were BF, white blood cell, age, TSH, TG, and LDL-C. We concluded that FPGbase was the most important factor to predict future prediabetes. However, after removing FPGbase, WBC, TSH, BF, HDL-C, and age were the key factors after 5.8 years.

Comparison between linear regression and four different machine learning methods in selecting risk factors for osteoporosis in a Chinese female aged cohort.

BACKGROUND: Population aging is emerging as an increasingly acute challenge for countries around the world. One particular manifestation of this phenomenon is the impact of osteoporosis on individuals and national health systems. Previous studies of risk factors for osteoporosis were conducted using traditional statistical methods, but more recent efforts have turned to machine learning approaches. Most such efforts, however, treat the target variable (bone mineral density [BMD] or fracture rate) as a categorical one, which provides no quantitative information. The present study uses five different machine learning methods to analyze the risk factors for T-score of BMD, seeking to (1) compare the prediction accuracy between different machine learning methods and traditional multiple linear regression (MLR) and (2) rank the importance of 25 different risk factors. METHODS: The study sample includes 24 412 women older than 55 years with 25 related variables, applying traditional MLR and five different machine learning methods: classification and regression tree, Naïve Bayes, random forest, stochastic gradient boosting, and eXtreme gradient boosting. The metrics used for model performance comparisons are the symmetric mean absolute percentage error, relative absolute error, root relative squared error, and root mean squared error. RESULTS: Machine learning approaches outperformed MLR for all four prediction errors. The average importance ranking of each factor generated by the machine learning methods indicates that age is the most important factor determining T-score, followed by estimated glomerular filtration rate (eGFR), body mass index (BMI), uric acid (UA), and education level. CONCLUSION: In a group of women older than 55 years, we demonstrated that machine learning methods provide superior performance in estimating T-Score, with age being the most important impact factor, followed by eGFR, BMI, UA, and education level.

Accuracy of Self-Injection Locking Radar System for Vital Signs Detection During the COVID-19 Pandemic at a Hospital in Taiwan: Measuring Vital Signs Accurately with SIL Radar for Hospital Healthcare.

BACKGROUND Self-injection locking (SIL) radar uses continuous-wave radar and an injection-locked oscillator-based frequency discriminator that receives and demodulates radar signals remotely to monitor vital signs. This study aimed to compare SIL radar with traditional electrocardiogram (ECG) measurements to monitor respiratory rate (RR) and heartbeat rate (HR) during the COVID-19 pandemic at a single hospital in Taiwan. MATERIAL AND METHODS We recruited 31 hospital staff members (16 males and 15 females) for respiratory rates (RR) and heartbeat rates (HR) detection. Data acquisition with the SIL radar and traditional ECG was performed simultaneously, and the accuracy of the measurements was evaluated using Bland-Altman analysis. RESULTS To analyze the results, participates were divided into 2 groups (individual subject and multiple subjects) by gender (male and female), or 4 groups (underweight, normal weight, overweight, and obesity) by body mass index (BMI). The results were analyzed using mean bias errors (MBE) and limits of agreement (LOA) with a 95% confidence interval. Bland-Altman plots were utilized to illustrate the difference between the SIL radar and ECG monitor. In all BMI groups, results of RR were more accurate than HR, with a smaller MBE. Furthermore, RR and HR measurements of the male groups were more accurate than those of the female groups. CONCLUSIONS We demonstrated that non-contact SIL radar could be used to accurately measure HR and RR for hospital healthcare during the COVID-19 pandemic.

Slow skeletal muscle troponin T acts as a potential prognostic biomarker and therapeutic target for hepatocellular carcinoma.

Slow skeletal muscle troponin T (TNNT1) as a poor prognostic indicator is upregulated in colon and breast cancers. However, the role of TNNT1 in the disease prognosis and biological functions of hepatocellular carcinoma (HCC) is still unclear. The Cancer Genome Atlas (TCGA), real-time quantitative RT-PCR (qRT-PCR), immunoblot, and immunohistochemical analyses were applied to evaluate the TNNT1 expression of human HCC. The impact of TNNT1 levels on disease progression and survival outcome was studied using TCGA analysis. Moreover, the bioinformatics analysis and HCC cell culture were used to investigate the biological functions of TNNT1. Besides, the immunoblot analysis and enzyme-linked immunosorbent assay (ELISA) were used to detect the extracellular TNNT1 of HCC cells and circulating TNNT1 of HCC patients, respectively. The effect of TNNT1 neutralization on oncogenic behaviors and signaling was further validated in the cultured hepatoma cells. In this study, tumoral and blood TNNT1 was upregulated in HCC patients based on the analyses using bioinformatics, fresh tissues, paraffin sections, and serum. From the multiple bioinformatics tools, the TNNT1 overexpression was associated with advanced stage, high grade, metastasis, vascular invasion, recurrence, and poor survival outcome in HCC patients. By the cell culture and TCGA analyses, TNNT1 expression and release were positively correlated with epithelial-mesenchymal transition (EMT) processes in HCC tissues and cells. Moreover, TNNT1 neutralization suppressed oncogenic behaviors and EMT in hepatoma cells. In conclusion, TNNT1 may serve as a non-invasive biomarker and drug target for HCC management. This research finding may provide a new insight for HCC diagnosis and treatment.

Long-term DEHP/MEHP exposure promotes colorectal cancer stemness associated with glycosylation alterations.

Plasticizers are considered as environmental pollution released from medical devices and increased potential oncogenic risks in clinical therapy. Our previous studies have shown that long-term exposure to di-ethylhexyl phthalate (DEHP)/mono-ethylhexyl phthalate (MEHP) promotes chemotherapeutic drug resistance in colorectal cancer. In this study, we investigated the alteration of glycosylation in colorectal cancer following long-term plasticizers exposure. First, we determined the profiles of cell surface N-glycomes by using mass spectrometry and found out the alterations of α2,8-linkages glycans. Next, we analyzed the correlation between serum DEHP/MEHP levels and ST8SIA6 expression from matched tissues in total 110 colorectal cancer patients. Moreover, clinical specimens and TCGA database were used to analyze the expression of ST8SIA6 in advanced stage of cancer. Finally, we showed that ST8SIA6 regulated stemness in vitro and in vivo. Our results revealed long-term DEHP/MEHP exposure significantly caused cancer patients with poorer survival outcome and attenuated the expression of ST8SIA6 in cancer cells and tissue samples. As expected, silencing of ST8SIA6 promoted cancer stemness and tumorigenicity by upregulating stemness-associated proteins. In addition, the cell viability assay showed enhanced drug resistance in ST8SIA6 silencing cells treated with irinotecan. Besides, ST8SIA6 was downregulated in the advanced stage and positively correlated with tumor recurrence in colorectal cancer. Our results imply that ST8SIA6 potentially plays an important role in oncogenic effects with long-term phthalates exposure.

Tendon Disorders in Chronic Liver Disease: A Retrospective Cohort Study in Taiwan.

To investigate the relationship between chronic liver disease and tendon disorder, a retrospective cohort study was conducted using the Kaohsiung Veterans General Hospital database. Patients >18 years with newly diagnosed liver disease and with at least a two-year follow-up in the hospital were included. An equal number of 20,479 cases were enrolled in both the liver-disease and non-liver-disease groups using a propensity score matching method. Disease was defined using ICD-9 or ICD-10 codes. The primary outcome was the development of tendon disorder. Demographic characteristics, comorbidities, use of tendon-toxic drugs, and status of HBV/HCV infection were included for analysis. The results showed 348 (1.7%) and 219 (1.1%) individuals developed tendon disorder in the chronic liver disease group and non-liver-disease group. Concomitant use of glucocorticoids and statins may have further raised the risk of tendon disorder in the liver disease group. The co-existence of HBV/HCV infection did not increase the risk of tendon disorder in the patients with liver disease. Considering these findings, physicians should be more aware of tendon issues in advance, and a prophylactic strategy should be adopted in patients with chronic liver disease.

A Positive Correlation between Steroid Injections and Cuff Tendon Tears: A Cohort Study Using a Clinical Database.

This cohort study aimed to investigate the association between steroid injections for shoulder diseases and the increased incidence of cuff tendon tears. The Kaohsiung Veterans General Hospital clinical database was used in this study. Patients were enrolled using the corresponding diagnostic codes for shoulder diseases. Patients who received steroid injections were included in the case group, and those without steroid injections were included in the control group. The outcome measure was the occurrence of cuff tendon tears during the study period. Adjusted hazard ratios for outcomes were calculated using Cox regression analysis adjusted for sex, age, and comorbidities. Of the 1025 patients with shoulder disease, 205 were in the case group and 820 were in the control group. The incidence of cuff tendon tears was 9.8% in patients who received steroid injections (p < 0.001). The adjusted hazard ratios for steroid injections, smoking, and chronic liver disease were 7.44 (p < 0.001), 2.40 (p = 0.046), 3.25 (p = 0.007), respectively. Steroid injections on the shoulder were associated with a raised risk of cuff tendon tears by 7.44 times compared to non-injection. The incidence of cuff tendon tears increased by 3.25 times with concurrent chronic liver disease and by 2.4 times with smoking.

DLK2 Acts as a Potential Prognostic Biomarker for Clear Cell Renal Cell Carcinoma Based on Bioinformatics Analysis.

Clear cell renal cell carcinoma (ccRCC) is the most common RCC subtype with a high mortality. It has been reported that delta-like 1 homologue (DLK1) participates in the tumor microenvironmental remodeling of ccRCC, but the relationship between delta-like 2 homologue (DLK2, a DLK1 homologue) and ccRCC is still unclear. Thus, this study aims to investigate the role of DLK2 in the biological function and disease prognosis of ccRCC using bioinformatics analysis. The TNMplot database showed that DLK2 was upregulated in ccRCC tissues. From the UALCAN analysis, the overexpression of DLK2 was associated with advanced stage and high grade in ccRCC. Moreover, the Kaplan-Meier plotter (KM Plotter) database showed that DLK2 upregulation was associated with poor survival outcome in ccRCC. By the LinkedOmics analysis, DLK2 signaling may participated in the modulation of ccRCC extracellular matrix (ECM), cell metabolism, ribosome biogenesis, TGF-ß signaling and Notch pathway. Besides, Tumor Immune Estimation Resource (TIMER) analysis showed that the macrophage and CD8+ T cell infiltrations were associated with good prognosis in ccRCC patients. Finally, DLK2 overexpression was associated with the reduced macrophage recruitments and the M1-M2 polarization of macrophage in ccRCC tissues. Together, DLK2 may acts as a novel biomarker, even therapeutic target in ccRCC. However, this study lacks experimental validation, and further studies are required to support this viewpoint.

Bioinformatics Analyses Identify the Therapeutic Potential of ST8SIA6 for Colon Cancer.

Sialylation of glycoproteins is modified by distinct sialyltransferases such as ST3Gal, ST6Gal, ST6GalNAc, or ST8SIA with α2,3-, α2,6-, or α2,8-linkages. Alteration of these sialyltransferases causing aberrant sialylation is associated with the progression of colon cancer. However, among the ST8- sialyltransferases, the role of ST8SIA6 in colon cancer remains poorly understood. In this study, we explored the involvement of ST8SIA6 in colon cancer using multiple gene databases. The relationship between ST8SIA6 expression and tumor stages/grades was investigated by UALCAN analysis, and Kaplan-Meier Plotter analysis was used to analyze the expression of ST8SIA6 on the survival outcome of colon cancer patients. Moreover, the biological functions of ST8SIA6 in colon cancer were explored using LinkedOmics and cancer cell metabolism gene DB. Finally, TIMER and TISMO analyses were used to delineate ST8SIA6 levels in tumor immunity and immunotherapy responses, respectively. ST8SIA6 downregulation was associated with an advanced stage and poorly differentiated grade; however, ST8SIA6 expression did not affect the survival outcomes in patients with colon cancer. Gene ontology analysis suggested that ST8SIA6 participates in cell surface adhesion, angiogenesis, and membrane vesicle trafficking. In addition, ST8SIA6 levels affected immunocyte infiltration and immunotherapy responses in colon cancer. Collectively, these results suggest that ST8SIA6 may serve as a novel therapeutic target towards personalized medicine for colon cancer.

Atrial Fibrillation Risk and Urate-Lowering Therapy in Patients with Gout: A Cohort Study Using a Clinical Database.

Individuals of Asian descent are at higher risk for developing hyperuricemia and gout as compared to Western populations. Urate-lowering therapy (ULT) is an effective treatment for hyperuricemia and gout. It was reported that febuxostat, one of the ULTs, raises the risk of atrial fibrillation (AF) in elderly populations. Nevertheless, this association has not been properly investigated in Asian populations. We aimed to investigate the development of AF after ULT with different drugs in an Asian population. We conducted a retrospective cohort study using the clinical database at Kaohsiung Veterans General Hospital. Patients newly diagnosed with gout between 1 January 2013 and 31 December 2020 and with a documented baseline serum uric acid (sUA) level but no prior diagnosis of AF were identified. Patients were divided into three groups-allopurinol, benzbromarone, and febuxostat users. During the follow-up period, the risks of incident AF following the initiation of ULT with different drugs were assessed. Development of incident AF was noted in 43 (6%) of the 713 eligible patients during the follow-up period (mean, 49.4 ± 26.6 months). Febuxostat-treated patients had a higher prevalence of certain comorbidities (diabetes mellitus, heart failure, and chronic kidney disease) and higher CHA2DS2-VASc scores. Compared with allopurinol, neither febuxostat nor benzbromarone was associated with increased adjusted hazard ratios (HR) for incident AF (HR: 1.20, 95% confidence interval [CI]: 0.43-3.34; HR: 0.68, 95% CI: 0.22-2.08). There was no difference in the risk of incident AF among Asian patients with gout who received febuxostat, allopurinol, or benzbromarone. Further studies are needed to evaluate long-term cardiovascular outcomes in patients receiving different ULT drugs.

The Effect of Practice toward Do-Not-Resuscitate among Taiwanese Nursing Staff Using Path Modeling.

This study aimed to elucidate the predictors and the effects of path modeling on the knowledge, attitude, and practice toward do-not-resuscitate (DNR) among the Taiwanese nursing staff. This study was a cross-sectional, descriptive design using stratified cluster sampling. We collected data on demographics, knowledge, attitude, and practice as measured by the DNR inventory (KAP-DNR), Mindful Attention Awareness Scale, General Self-Efficacy Scale, and Dispositional Resilience Scale. Participants were 194 nursing staff from a medical center in northern Taiwan in 2019. The results showed that participation in DNR signature and education related to palliative care were significant positive predictors of knowledge toward DNR. The DNR predictors toward attitude included DNR knowledge, mindfulness, self-efficacy, dispositional resilience, and religious belief of nurses. Generally, the critical predictors of DNR practice were DNR attitude, dispositional resilience, and male nurses. In path modeling, we identified that self-efficacy, dispositional resilience, master's degree, and religious belief directly influenced practice constituting DNR. Based on the findings of this study, we propose that nurses should improve their self-efficacy and dispositional resilience through training programs. Encouraging staff to undertake further education and have religious beliefs can enhance the practice of DNR and provide better end-of-life care.

TP3, an antimicrobial peptide, inhibits infiltration and motility of glioblastoma cells via modulating the tumor microenvironment.

Glioblastoma multiforme (GBM) is a cancer of the central nervous system with limited therapeutic outcomes. Infiltrating cancer cells are the contributing factor to high GBM malignancy. The intracranial brain cancer cell infiltration is a complex cascade involving adhesion, migration, and invasion. An arsenal of natural products has been under exploration to overcome GBM malignancy. This study applied the antimicrobial peptide tilapia piscidin 3 (TP3) to GBM8401, U87MG, and T98G cells. The cellular assays and microscopic observations showed that TP3 significantly attenuated cell adhesion, migration, and invasion. A live-cell video clip showed the inhibition of filopodia protrusions and cell attachment. Probing at the molecular levels showed that the proteolytic activities (from secretion), the mRNA and protein expression levels of matrix metalloproteinases-2 and -9 were attenuated. This result strongly evidenced that both invasion and metastasis were inhibited, although metastatic GBM is rare. Furthermore, the protein expression levels of cell-mobilization regulators focal adhesion kinase and paxillin were decreased. Similar effects were observed in small GTPase (RAS), phosphorylated protein kinase B (AKT) and MAP kinases such as extracellular signal-regulated kinases (ERK), JNK, and p38. Overall, TP3 showed promising activities to prevent cell infiltration and metastasis through modulating the tumor microenvironment balance, suggesting that TP3 merits further development for use in GBM treatments.