Search

BackgroundTo address gaps in global understanding of cultural and social variations, this study used a high-performance machine learning (ML) model to predict adolescent substance use across three national datasets. ObjectiveThis study aims to develop a generalizable predictive model for adolescent substance use using multinational datasets and ML. MethodsThe study used the Korea Youth Risk Behavior Web-Based Survey (KYRBS) from South Korea (n=1,098,641) to train ML models. For external validation, we used the Youth Risk Behavior Survey (YRBS) from the United States (n=2,511,916) and Norwegian nationwide Ungdata surveys (Ungdata) from Norway (n=700,660). After developing various ML models, we evaluated the final model’s performance using multiple metrics. We also assessed feature importance using traditional methods and further analyzed variable contributions through SHapley Additive exPlanation values. ResultsThe study used nationwide adolescent datasets for ML model development and validation, analyzing data from 1,098,641 KYRBS adolescents, 2,511,916 YRBS participants, and 700,660 from Ungdata. The XGBoost model was the top performer on the KYRBS, achieving an area under receiver operating characteristic curve (AUROC) score of 80.61% (95% CI 79.63-81.59) and precision of 30.42 (95% CI 28.65-32.16) with detailed analysis on sensitivity of 31.30 (95% CI 29.47-33.20), specificity of 99.16 (95% CI 99.12-99.20), accuracy of 98.36 (95% CI 98.31-98.42), balanced accuracy of 65.23 (95% CI 64.31-66.17), F1-score of 30.85 (95% CI 29.25-32.51), and area under precision-recall curve of 32.14 (95% CI 30.34-33.95). The model achieved an AUROC score of 79.30% and a precision of 68.37% on the YRBS dataset, while in external validation using the Ungdata dataset, it recorded an AUROC score of 76.39% and a precision of 12.74%. Feature importance and SHapley Additive exPlanation value analyses identified smoking status, BMI, suicidal ideation, alcohol consumption, and feelings of sadness and despair as key contributors to the risk of substance use, with smoking status emerging as the most influential factor. ConclusionsBased on multinational datasets from South Korea, the United States, and Norway, this study shows the potential of ML models, particularly the XGBoost model, in predicting adolescent substance use. These findings provide a solid basis for future research exploring additional influencing factors or developing targeted intervention strategies.

Abstract Previous studies have examined the prevalence of allergic diseases in adolescents 1–2 years after the emergence of the COVID-19 pandemic. However, more data is needed to understand the long-term impact of COVID-19 on allergic diseases. Thus, we aimed to examine the trend of the atopic dermatitis prevalence in Korean adolescents before and during the COVID-19 pandemic across 14 years. Additionally, we analyze the risk factors of atopic dermatitis (AD) based on the results. The Korean Disease Control and Prevention Agency conducted the Korea Youth Risk Behavior Web-based Survey from 2009 to 2022, from which the data for this study were obtained. Prevalence trends were compared across subgroups, and the β difference (βdiff) was calculated. We computed odds ratios to examine changes in the disease prevalence before and during the pandemic. This study included a total of 917,461 participants from 2009 to 2022. The prevalence of atopic dermatitis increased from 6.79% (95% CI 6.66–6.91) in 2009–2011 to 6.89% (95% CI 6.72–7.05) in 2018–2019, then decreased slightly to 5.82% (95% CI 5.60–6.04) in 2022. Across the 14 years, middle school student status, low parent’s highest education level, low household income, non-alcohol consumption, non-smoker smoking status, no suicidal thoughts, and no suicide attempts were associated with increased risk of atopic dermatitis, while female sex, rural residence, high BMI, low school performance, low household income, and no feelings of sadness and despair was associated with a small increase. This study examined the prevalence of atopic dermatitis across an 18-year, and found that the prevalence increased in the pre-pandemic then decreased during the start of the pandemic and remained constant throughout the pandemic. This trend could be explained mainly by the large scale social and political changes that occurred during the COVID-19 pandemic.

BackgroundUnderstanding the association between hypertension prevalence and socioeconomic and behavioral variables during a pandemic is essential, and this analysis should extend beyond short-term trends. ObjectiveThis study aims to examine long-term trends in the prevalence of participants diagnosed with and receiving treatment for hypertension, using data collected by a nationally representative survey from 2009 to 2022, which includes the COVID-19 pandemic era. MethodsA nationwide, population-based, cross-sectional study used data collected from the South Korea Community Health Survey between 2009 and 2022. The study sample comprised 3,208,710 Korean adults over a period of 14 years. We aimed to assess trends in the prevalence of participants diagnosed with and receiving treatment for hypertension in the national population from 2009 to 2022, with a specific focus on the COVID-19 pandemic, using weighted linear regression models. ResultsAmong the included 3,072,546 Korean adults, 794,239 (25.85%) were aged 19-39 years, 1,179,388 (38.38%) were aged 40-59 years; 948,097 (30.86%) were aged 60-79 years, and 150,822 (4.91%) were aged 80 years or older. A total of 1,426,379 (46.42%) were men; 761,896 (24.80%) and 712,264 (23.18%) were diagnosed with and received treatment for hypertension, respectively. Although the overall prevalence over the 14-year period increased, the upward trends of patients diagnosed with and receiving treatment for hypertension decreased during the COVID-19 pandemic era compared with the prepandemic era (β difference for trend during vs before the pandemic –.101, 95% CI –0.107 to –0.094 vs –.133, 95% CI –0.140 to –0.127). Notably, the trends in prevalence during the pandemic were less pronounced in subgroups of older adults (≥60 years old) and individuals with higher alcohol consumption (≥5 days/month). ConclusionsThis nationwide representative study found that the national prevalence of participants diagnosed with and receiving treatment for hypertension increased during the prepandemic era. However, there was a marked decrease in these trends during the prepandemic era, compared with the pandemic era, particularly among specific subgroups at increased risk of negative outcomes. Future studies are needed to evaluate the factors associated with changes in the prevalence of hypertension during the COVID-19 pandemic.

Background: Intrahepatic cholangiocarcinoma (IHCC) is highly aggressive primary hepatic malignancy with an increasing incidence. Objective: This study aimed to develop machine learning-based radiomic clustering using F-18 fluorodeoxyglucose (FDG) positron emission tomography/computed tomography (PET/CT) for predicting recurrence-free survival (RFS) and overall survival (OS) in IHCC. Methods: We retrospectively reviewed pretreatment F-18 FDG PET/CT scans of 60 IHCC patients who underwent surgery without neoadjuvant treatment between January 2008 and July 2020. Radiomic features such as first order, shape, and gray level were extracted from the scans of 52 patients and analyzed using unsupervised hierarchical clustering. Results: Of the 60 patients, 36 experienced recurrence and 31 died during follow-up. Eight patients with a negative FDG uptake were classified as Group 0. The unsupervised hierarchical clustering analysis divided the total cohort into three clusters (Group 1: n = 27; Group 2: n = 23; Group 3: n = 2). The Kaplan–Meier curves showed significant differences in RFS and OS among the clusters (p < 0.0001). Multivariate analyses showed that the PET radiomics grouping was an independent prognostic factor for RFS (hazard ratio (HR) = 3.03, p = 0.001) and OS (HR = 2.39, p = 0.030). Oxidative phosphorylation was significantly activated in Group 1, and the KRAS, P53, and WNT β-catenin pathways were enriched in Group 2. Conclusions: This study demonstrated that machine learning-based PET radiomics clustering can preoperatively predict prognosis and provide valuable information complementing the genomic profiling of IHCC.

Abstract The impact of the pandemic on hypertension management is unknown, particularly regarding changes in demographic risk factors. We conducted a comprehensive study between 1998 and 2021 on the long-term trends in hypertension prevalence in South Korea, including a comparison of the pre-pandemic and pandemic eras. Data from 1998 to 2021 of 108,687 Korean adults were obtained through a nationwide, large-scale, and serial study. We conducted a weighted complex sampling analysis on the estimates of national prevalence and compared the slope of hypertension prevalence before and during the pandemic to determine the trend dynamics. We included 108,687 participants over 24 years, 1998–2021. While the prevalence of patients with hypertension consistently increased before the pandemic from 25.51% [95% CI: 24.27–26.75] in 1998–2005 to 27.81% [95% CI: 26.97–28.66] in 2016–2019, the increasing slope in hypertension prevalence slowed during the pandemic period (28.07% [95% CI: 26.16–29.98] for 2021; βdiff, −0.012 [−0.023 to 0.000]). Hypertension awareness, treatment, control, and control rates among patients receiving treatment followed similar trends. Compared to the pre-pandemic era, individuals aged 19–59 years or male had significantly increased control rates among the treated patients during the pandemic. This study investigated long-term trends in hypertension prevalence, awareness, treatment, and control among Korean adults. The absence of a reduction in the health indicators associated with hypertension during the pandemic implies that medical services for individuals with hypertension remain unaffected.

Abstract Studies on the trends in the prevalence of rheumatoid arthritis (RA) and osteoarthritis (OA) are limited, particularly during the COVID-19 pandemic. This study aimed to analyze the temporal trend of RA and OA in South Korean adults from 1998 to 2021, including the COVID-19 pandemic period. The Korea National Health and Nutrition Examination Survey (KNHANES) data on adults aged ≥ 19 years were analyzed to investigate the prevalence of RA and OA from 1998 to 2021. The prevalence trends were compared by the years, and βdiff (β difference) was calculated. Odds ratios (ORs) were computed for each disease to examine changes in disease prevalence before and during the pandemic in order to determine the impact of the pandemic on disease prevalence. Among 163,221 Korean adults, the prevalence of RA and OA showed a steady decrease from 2005 (RA: from 1.91% in 2005–2007 to 1.55% in 2016–2019 and OA: from 9.75% in 2005–2007 to 8.27% in 2016–2019), but there was a slight increased after the onset of the COVID-19 pandemic (RA: from 1.23% in 2020 to 1.36% in 2021 and OA: from 8.04% in 2020 to 8.27% in 2021). Vulnerable groups, including participants aged ≥ 60 years (versus 19–60 years, ratio of ORs: 1.222; 95% CI 1.011–1.477), urban residents (ratio of ORs: 1.289; 95% CI 1.007–1.650), and participants with higher education level (ratio of ORs: 1.360; 95% CI 1.119–1.653) showed higher ORs of OA, whereas no particularly vulnerable population was observed for RA. Our findings provide an insight into the long-term trends of RA and OA among adult population and highlight a novel perspective on the impact of COVID-19 on disease prevalence.

BackgroundSuicide is the second-leading cause of death among adolescents and is associated with clusters of suicides. Despite numerous studies on this preventable cause of death, the focus has primarily been on single nations and traditional statistical methods. ObjectiveThis study aims to develop a predictive model for adolescent suicidal thinking using multinational data sets and machine learning (ML). MethodsWe used data from the Korea Youth Risk Behavior Web-based Survey with 566,875 adolescents aged between 13 and 18 years and conducted external validation using the Youth Risk Behavior Survey with 103,874 adolescents and Norway’s University National General Survey with 19,574 adolescents. Several tree-based ML models were developed, and feature importance and Shapley additive explanations values were analyzed to identify risk factors for adolescent suicidal thinking. ResultsWhen trained on the Korea Youth Risk Behavior Web-based Survey data from South Korea with a 95% CI, the XGBoost model reported an area under the receiver operating characteristic (AUROC) curve of 90.06% (95% CI 89.97-90.16), displaying superior performance compared to other models. For external validation using the Youth Risk Behavior Survey data from the United States and the University National General Survey from Norway, the XGBoost model achieved AUROCs of 83.09% and 81.27%, respectively. Across all data sets, XGBoost consistently outperformed the other models with the highest AUROC score, and was selected as the optimal model. In terms of predictors of suicidal thinking, feelings of sadness and despair were the most influential, accounting for 57.4% of the impact, followed by stress status at 19.8%. This was followed by age (5.7%), household income (4%), academic achievement (3.4%), sex (2.1%), and others, which contributed less than 2% each. ConclusionsThis study used ML by integrating diverse data sets from 3 countries to address adolescent suicide. The findings highlight the important role of emotional health indicators in predicting suicidal thinking among adolescents. Specifically, sadness and despair were identified as the most significant predictors, followed by stressful conditions and age. These findings emphasize the critical need for early diagnosis and prevention of mental health issues during adolescence.

Abstract Background We aimed to quantify the identification of mild cognitive impairment and/or Alzheimer’s disease using olfactory-stimulated functional near-infrared spectroscopy using machine learning through a post hoc analysis of a previous diagnostic trial and an external additional trial. Methods We conducted two independent, patient-level, single-group, diagnostic interventional trials (original and additional trials) involving elderly volunteers (aged > 60 years) with suspected declining cognitive function. All volunteers were assessed by measuring the oxygenation difference in the orbitofrontal cortex using an open-label olfactory-stimulated functional near-infrared spectroscopy approach, medical interview, amyloid positron emission tomography, brain magnetic resonance imaging, Mini-Mental State Examination, and Seoul Neuropsychological Screening Battery. Results In total, 97 (original trial) and 36 (additional trial) elderly volunteers with suspected decline in cognitive function met the eligibility criteria. The statistical model reported classification accuracies of 87.3% in patients with mild cognitive impairment and Alzheimer’s disease in internal validation (original trial) but 63.9% in external validation (additional trial). The machine learning algorithm achieved 92.5% accuracy with the internal validation data and 82.5% accuracy with the external validation data. For the diagnosis of mild cognitive impairment, machine learning performed better than statistical methods with internal (86.0% versus 85.2%) and external validation data (85.4% versus 68.8%). Interpretation In two independent trials, machine learning models using olfactory-stimulated oxygenation differences in the orbitofrontal cortex were superior in diagnosing mild cognitive impairment and Alzheimer’s disease compared to classic statistical models. Our results suggest that the machine learning algorithm is stable across different patient groups and increases generalization and reproducibility. Trial registration Clinical Research Information Service (CRiS) of Republic of Korea; CRIS numbers, KCT0006197 and KCT0007589. Graphical Abstract

Summary: Background: Autoimmune hepatitis (AIH) varies significantly in incidence and prevalence across countries and regions. We aimed to examine global, regional, and national trends in incidence and prevalence of AIH from 1970 to 2022. Methods: We conducted a thorough search of the PubMed/MEDLINE, Embase, CINAHL, Google Scholar, and Cochrane databases from database inception to August 9, 2023, using the search term “autoimmune hepatitis” in combination with “incidence,” “prevalence,” or “trend.” Only general population-based observational studies with larger samples sizes were considered for inclusion. Studies that recruited convenience samples, and those with fewer than 50 participants were excluded. Summary data were extracted from published reports. A random effects model was used and pooled estimates with 95% CI were used to calculate the incidence and prevalence of AIH. Heterogeneity was evaluated using the I2 statistic. The study protocol was registered with PROSPERO, CRD42023430138. Findings: A total of 37 eligible studies, encompassing more than 239 million participants and 55,839 patients with AIH from 18 countries across five continents, were included in the analysis. Global pooled incidence and prevalence of AIH were found to be 1.28 cases per 100,000 inhabitant-years (95% CI, 1.01–1.63, I2 = 99·51%; number of studies, 33; sample population, 220,673,674) and 15.65 cases per 100,000 inhabitants (95% CI, 13.42–18.24, I2 = 99·75%; number of studies, 26; sample population, 217,178,684), respectively. The incidence of AIH was greater in countries with high Human Development Index (>0.92), in North America and Oceania (compared with Asia), among females, adults (compared with children), and high latitude (>45°). Similar patterns in AIH prevalence were observed. Pooled AIH prevalence increased gradually from 1970 to 2019 (1970–1999; 9.95 [4.77–15.13], I2 = 95·58% versus 2015–2022; 27.91 [24.86–30.96], I2 = 99·32%; cases per 100,000 inhabitants). The overall incidence and prevalence of AIH, as well as some subgroup analyses of the studies, displayed asymmetry in the funnel plots, suggesting potential evidence of publication bias. Interpretation: AIH incidence and prevalence have increased significantly and exhibit substantial variation across regions worldwide. Further research is required to assess the incidence and prevalence of AIH, specifically in South America and Africa. Funding: National Research Foundation of Korea.

Objective: Famotidine has been proposed as a promising candidate for the treatment of coronavirus disease 2019 (COVID-19). However, there is limited research on the association of famotidine with the poor prognosis of COVID-19. Methods: The Korean nationwide cohort included 6,556 patients who tested positive on RT-PCR for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). The poor COVID-19-related outcomes were defined on the basis of having encountered the composite outcome of high oxygen therapy, intensive care unit admission, administration of mechanical ventilation, or death. In addition, we performed exposure-driven propensity score matching for no H2-blocker use versus current famotidine use, and other H2-blocker use versus current famotidine use. Results: 4,785 (73.0%) patients did not use a H2-blocker, 393 (6.0%) patients were currently used famotidine, and 1,292 (19.7%) patients currently used H2-blocker other than famotidine. In multivariable analysis after matching (no H2-blocker use versus current famotidine use), there was no significant association between current famotidine use and composite outcomes (adjusted odd ratios [aOR]: 1.30, 95% confidence interval [CI]: 0.55–3.06). On the other hand, another matched cohort (other H2-blocker use versus current famotidine use), demonstrated a positive association between current famotidine use and composite outcomes (aOR: 3.56, 95% CI: 1.03–12.28) Conclusions: Our study results did not support the potential of famotidine as a therapeutic agent for COVID-19. A rather unexpected result could be observed in the comparisons between current famotidine use and other H2-blocker use; it was observed that current famotidine use increased the risk of poor COVID-19 related outcomes. Further studies are needed to clearly prove the causal relationship with several H2-blockers, including famotidine.