Epidemiology of pediatric inflammatory bowel disease categorized by age subgroups in Korea.

BACKGROUND: Pediatric inflammatory bowel disease (PIBD) affects different age groups and its incidence is increasing worldwide. However, there is a lack of research focusing on age subgroups in Asian countries. In this nationwide population-based study, we investigated the epidemiology of PIBD among different age subgroups in Korea. METHODS: We analyzed Korean health administration data from 2005 to 2016. Data were divided by age at diagnosis as follows: group 1, 0-1 years; group 2, 2-5 years; group 3, 6-9 years; group 4, 10-16 years. We analyzed the overall incidence, temporal changes, and regional differences by age subgroups, using Poisson regression analysis. RESULTS: From 2005 to 2016, 2734 inflammatory bowel disease (IBD) cases were diagnosed among patients under 17 years of age. In the overall population, the incidence rate of PIBD over the entire study period was 2.248/105 person-years (PY), significantly increasing from 1.173/105 PY in 2005-2007 to 3.267/105 PY in 2014-2016. The incidence rates in groups 1 and 2 remained unchanged, whereas those of groups 3 and 4 increased significantly. The same trend was observed when analyzed separately for Crohn's disease (CD) and ulcerative colitis (UC). The incidence rates of CD in groups 3 and 4 showed differences between metropolitan and non-metropolitan areas, whereas those in groups 1 and 2, and UC of all age subgroups showed no difference. CONCLUSIONS: The temporal trend and regional differences of PIBD differed among age subgroups, suggesting that genetic and environmental factors have varying impacts on IBD development across different subgroups.

Continual learning framework for a multicenter study with an application to electrocardiogram.

Deep learning has been increasingly utilized in the medical field and achieved many goals. Since the size of data dominates the performance of deep learning, several medical institutions are conducting joint research to obtain as much data as possible. However, sharing data is usually prohibited owing to the risk of privacy invasion. Federated learning is a reasonable idea to train distributed multicenter data without direct access; however, a central server to merge and distribute models is needed, which is expensive and hardly approved due to various legal regulations. This paper proposes a continual learning framework for a multicenter study, which does not require a central server and can prevent catastrophic forgetting of previously trained knowledge. The proposed framework contains the continual learning method selection process, assuming that a single method is not omnipotent for all involved datasets in a real-world setting and that there could be a proper method to be selected for specific data. We utilized the fake data based on a generative adversarial network to evaluate methods prospectively, not ex post facto. We used four independent electrocardiogram datasets for a multicenter study and trained the arrhythmia detection model. Our proposed framework was evaluated against supervised and federated learning methods, as well as finetuning approaches that do not include any regulation to preserve previous knowledge. Even without a central server and access to the past data, our framework achieved stable performance (AUROC 0.897) across all involved datasets, achieving comparable performance to federated learning (AUROC 0.901).

Association between Sight-Threatening Eye Diseases and Death by Suicide in South Korea: A Nationwide Population-based Cohort Study.

PURPOSE: Although associations between visual impairment (VI) and suicide are posited, specific risks across the sight-threatening eye disease (STED) spectrum remain to be assessed. We determined whether individuals with STED die more often by suicide than do other people and assessed the temporal associations. DESIGN: A nationwide, population-based cohort study. PARTICIPANTS: All persons aged 40 years or older in South Korea from 2010 to 2020. METHODS: Persons diagnosed with STEDs (i.e., glaucoma, exudative age-related macular degeneration [AMD], or diabetic retinopathy [DR]) were identified in the Korean National Health Insurance (NHI) service database. Both NHI health checkup records and the National Disability Registration were used for coexisting severe VI. Death by suicide was defined as diagnostic codes as recorded in the Korea National Statistical Office. Incidence rate ratios (IRRs) were estimated by quasi-Poisson regressions and adjusted for sociodemographics, comorbidity, psychiatric diagnoses, and VI. The temporal relationship between time since first STED diagnosis and suicide risk was determined by identifying patients with STED newly diagnosed during the period from 2010 to 2011. MAIN OUTCOME MEASURES: The IRR of death by suicide in people with STED relative to those without. RESULTS: Of the 2.8 million people (45% male) observed for 24 300 969 person-years, 13 205 died by suicide. Among them, 34% (n = 4514) had a STED diagnosis, for a suicide rate of 69 per 100 000 person-years (95% confidence interval [CI], 67-72), relative to 51 per 100 000 person-years (95% CI, 50-52) for non-STED individuals. People with STED had an adjusted IRR of 1.33 (95% CI, 1.26-1.41) relative to those without. The largest excess adjusted IRR of suicide mortality was that for DR (1.40, 95% CI, 1.29-1.52). For exudative AMD, the adjusted IRR was 1.20 (95% CI, 1.04-1.39), whereas for glaucoma, the corresponding value was 1.09 (95% CI, 1.02-1.17). With coexisting severe VI, the IRR for any STED was 1.49 (95% CI, 1.29-1.73). The highest suicide hazard ratio was between 3 and 6 months postdiagnosis (5.33; 95% CI, 4.59-6.20). CONCLUSIONS: In South Korea between 2010 and 2020, a higher suicide rate was evident among those with diagnosed STED than for persons not so diagnosed. FINANCIAL DISCLOSURE(S): The author(s) have no proprietary or commercial interest in any materials discussed in this article.

Shift to a Younger Age and Regional Differences in Inflammatory Bowel Disease in Korea: Using Healthcare Administrative Data.

BACKGROUND: Research using healthcare administrative data with a validated algorithm can reveal the real-world data of rare diseases. AIMS: We investigated an accurate algorithm for detecting incident cases of inflammatory bowel disease (IBD) from healthcare data and analyzed the nationwide population-based epidemiological features in Korea. METHODS: Healthcare data from Songpa-Kangdong districts in Seoul were extracted from the National Health Insurance Service and analyzed to identify the best algorithm reflecting the cohort data. The most accurate criterion was applied to the entire database for further analysis. RESULTS: With the selected working criteria, 37,555 incident cases of IBD (Crohn's Disease [CD], 13,130; ulcerative colitis [UC], 24,425) were identified from 2005 to 2016. The male-to-female ratio was 2.5:1 for CD and 1.4:1 for UC. Over 12 years, the annual standardized incidence rate (SIR) per 100,000 people increased from 1.6 to 2.7 and 3.8 to 4.3 for CD and UC, respectively. The peak age at diagnosis of UC shifted from 55-59 years to 20-24 years, whereas that of CD shifted from 19 to 17 years. The SIR of CD was higher in metropolitan areas than in non-metropolitan areas. CONCLUSIONS: This nationwide population-based epidemiologic study of Korean IBD revealed a gradual increase in the incidence rates and a notable shift toward younger age at diagnosis. Males were predominant in both CD and UC. There was an urban-rural difference in the SIR of CD.

Association between higher variability in kidney function and long-term mortality.

AIM: We evaluated whether estimated glomerular filtration rate variability in the general population could be associated with all-cause mortality. METHODS: Health examination data from 7842 individuals aged >20 years who visited for health check-ups at least thrice at ≥6-month intervals between May 1, 1995 and November 30, 2010 were collected. Estimated glomerular filtration rate variability was defined as the coefficient of variation of the estimated glomerular filtration rate, that is, standard deviation/mean value multiplied by 100. The study population was divided into three groups based on the coefficient of variation tertiles, and the mortality risks were compared across groups. RESULTS: The mean duration from the final visit to the outcome was 10.3 ± 2.9 years. The mean coefficient of variations of estimated glomerular filtration rate variability from the lowest to the highest variability group were 5.1 ± 1.8%, 9.0 ± 1.0%, and 14.4 ± 3.9%, respectively. There was a 1.3 times higher risk of mortality in the group with the highest variability (hazard ratio: 1.300, 95% confidence interval: 1.013-1.669) after adjustment. The findings were similar in patients with diabetes and those >60 years old (hazard ratio: 1.635, 95% confidence interval: 1.076-2.483; hazard ratio: 1.585, 95% confidence interval: 1.107-2.269). CONCLUSION: Higher estimated glomerular filtration rate variability was associated with increased 10-year mortality in the general population. This variability was very small, but considering the patients' long-term prognoses, it was significant.

Simple Postoperative AKI Risk (SPARK) Classification before Noncardiac Surgery: A Prediction Index Development Study with External Validation.

BACKGROUND: Researchers have suggested models to predict the risk of postoperative AKI (PO-AKI), but an externally validated risk index that can be practically implemented before patients undergo noncardiac surgery is needed. METHODS: We performed a retrospective observational study of patients without preexisting renal failure who underwent a noncardiac operation (≥1 hour) at two tertiary hospitals in Korea. We fitted a proportional odds model for an ordinal outcome consisting of three categories: critical AKI (defined as Kidney Disease Improving Global Outcomes AKI stage ≥2, post-AKI death, or dialysis within 90 days after surgery), low-stage AKI (defined as PO-AKI events not fulfilling the definition of critical AKI), and no PO-AKI. RESULTS: The study included 51,041 patients in a discovery cohort and 39,764 patients in a validation cohort. The Simple Postoperative AKI Risk (SPARK) index included a summation of the integer scores of the following variables: age, sex, expected surgery duration, emergency operation, diabetes mellitus, use of renin-angiotensin-aldosterone inhibitors, baseline eGFR, dipstick albuminuria hypoalbuminemia, anemia, and hyponatremia. The model calibration plot showed tolerable distribution of observed and predicted probabilities in both cohorts. The discrimination power of the SPARK index was acceptable in both the discovery (c-statistic 0.80) and validation (c-statistic 0.72) cohorts. When four SPARK classes were defined on the basis of the sum of the risk scores, the SPARK index and classes fairly reflected the risks of PO-AKI and critical AKI. CONCLUSIONS: Clinicians may consider implementing the SPARK index and classifications to stratify patients' PO-AKI risks before performing noncardiac surgery.

Artificial Intelligence in Health Care: Current Applications and Issues.

In recent years, artificial intelligence (AI) technologies have greatly advanced and become a reality in many areas of our daily lives. In the health care field, numerous efforts are being made to implement the AI technology for practical medical treatments. With the rapid developments in machine learning algorithms and improvements in hardware performances, the AI technology is expected to play an important role in effectively analyzing and utilizing extensive amounts of health and medical data. However, the AI technology has various unique characteristics that are different from the existing health care technologies. Subsequently, there are a number of areas that need to be supplemented within the current health care system for the AI to be utilized more effectively and frequently in health care. In addition, the number of medical practitioners and public that accept AI in the health care is still low; moreover, there are various concerns regarding the safety and reliability of AI technology implementations. Therefore, this paper aims to introduce the current research and application status of AI technology in health care and discuss the issues that need to be resolved.

Acute Kidney Injury in Pediatric Cancer Patients.

OBJECTIVE: To analyze the incidence of acute kidney injury (AKI) in the first year after cancer diagnosis in children and to evaluate the short-term and long-term effects on renal function and proteinuria. STUDY DESIGN: Retrospective review of medical records was done on children who were diagnosed and treated for cancer at Seoul National University Hospital between 2004 and 2013. AKI was defined according to the Kidney Disease: Improving Global Outcomes criteria. Impaired renal function of estimated glomerular filtration rate less than 90 mL/minute/1.73 m2 and development of proteinuria of cancer survivors were also assessed. RESULTS: This study included 1868 patients who were diagnosed with cancer at a median age of 7.9 years. During the course of treatment, 983 patients (52.6%) developed 1864 episodes of AKI, and the cumulative incidence at 2 weeks, 3 months, and 1 year after diagnosis was 28.9%, 39.6%, and 53.6%, respectively. The 1-year cumulative incidence was the highest in patients with acute myeloid leukemias (88.4%). In all, 6.1% of patients had more than 4 episodes of AKI and 11.8% of patients had stage 3 AKI. Among the 1096 childhood cancer survivors, 22.6% were found to have impaired renal function. A greater number of AKI episodes (≥4 times) and nephrectomy were independent risk factors of impaired renal function. Also, 8.2% of the survivors developed proteinuria among 742 childhood cancer survivors. CONCLUSIONS: A large percentage of children with cancer experience AKI during the course of treatment, and AKI is associated with impaired long-term renal function.

Different association between renal hyperfiltration and mortality by sex.

AIM: Renal hyperfiltration (RHF) is a marker of early kidney injury that was recently shown to be a novel marker of mortality. However, it has no clear definition. In this study, we suggested an age- and sex-adjusted RHF definition and explored the association between RHF and mortality by sex. METHODS: We analyzed data from individuals receiving routine health examinations from 1995 to 2009. RHF was defined as an estimated glomerular filtration rate over the 95th percentile matched for age and sex. RESULTS: A total of 114 966 individuals were included. During the 75-month of observation period, 2559 (2.2%) participants died. Among those, 71.4% were men. Because sex and RHF had a significant interaction for mortality (P for interaction < 0.001), we performed survival analysis according to sex. RHF was related to lower body weight and a higher proportion of cigarette smoking in men, whereas these relationships were not found in women. In the Kaplan-Meier curve, RHF was associated with higher mortality rate than non-RHF in both sexes, but this relationship was more prominent in men. In the multivariate analysis, RHF remained as an independent risk factor for all-cause mortality even after adjustment for confounding in men (hazard ratio, 1.34; 95% confidence interval, 1.12-1.59; P = 0.001). In women, RHF was not associated with increased mortality. CONCLUSIONS: We demonstrated that RHF was a significant risk factor for mortality in men but not in women. The mechanisms and clinical implications of these different associations according to sex require a further clarification.

Unobtrusive Estimation of Cardiorespiratory Fitness with Daily Activity in Healthy Young Men.

Despite the importance of cardiorespiratory fitness, no practical method exists to estimate maximal oxygen consumption (VO2max) without a specific exercise protocol. We developed an estimation model of VO2max, using maximal activity energy expenditure (aEEmax) as a new feature to represent the level of physical activity. Electrocardiogram (ECG) and acceleration data were recorded for 4 days in 24 healthy young men, and reference VO2max levels were measured using the maximal exercise test. aEE was calculated using the measured acceleration data and body weight, while heart rate (HR) was extracted from the ECG signal. aEEmax was obtained using linear regression, with aEE and HR as input parameters. The VO2max was estimated from the aEEmax using multiple linear regression modeling in the training group (n = 16) and was verified in the test group (n = 8). High correlations between the estimated VO2max and the measured VO2max were identified in both groups, with a 15-hour recording being sufficient to produce a highly accurate VO2max estimate. Additional recording time did not significantly improve the accuracy of the estimation. Our VO2max estimation method provides a robust alternative to traditional approaches while only requiring minimal data acquisition time in daily life.

Mean corpuscular volume levels and all-cause and liver cancer mortality.

BACKGROUND: An elevated mean corpuscular volume (MCV) is associated with aging, nutrition, alcohol abuse and more, and it is known as a survival predictor in chronically ill patients. The aim of this study was to investigate the association between MCV levels and mortality from all-causes, cancer and site-specific cancer in a non-anemic healthy population. METHODS: A total of 36,260 participants aged 40 years or older who underwent routine check-ups at Seoul National University Hospital Health Promotion Center between 1995 and 2008 were followed-up for mortality until December 31, 2008, retrospectively. RESULTS: During an average follow-up of 8.0 years, 1107 deaths including 547 cancer deaths were observed. The adjusted hazard ratios (aHRs) of the subjects with the highest quartile of MCV ≥95.8 fL in men and MCV ≥94.2 fL in women for all-cause and cancer mortality were 1.44 [95% confidence interval (CI), 1.15-1.80] and 1.51 (95% CI, 1.10-2.07) for men and 1.55 (95% CI, 1.08-2.22) and 1.25 (95% CI, 0.74-2.11) for women, respectively, compared with those in the reference group (90.5 fL≤MC <93.0 fL in men and 89.2 fL≤MCV<91.6 fL in women). Elevated MCV level was related to an increased risk of liver cancer mortality in men (aHR, 3.55; 95% CI, 1.75-7.21). CONCLUSIONS: This study suggests that the elevated MCV level in non-anemic cancer-free individuals was associated with increased all-cause mortality in both men and women, and with cancer mortality, in particular liver cancer mortality in men. Future prospective studies are required to consolidate our findings.

Association between lower serum bicarbonate and renal hyperfiltration in the general population with preserved renal function: a cross-sectional study.

BACKGROUND: Lower serum bicarbonate, mainly due to the modern Western-style diet, and renal hyperfiltration (RHF) are both independently associated with higher mortality in the general population with preserved renal function. The objective of this study was to evaluate the association between serum bicarbonate and RHF. METHODS: The health data of 41,886 adults with an estimated glomerular filtration rate (eGFR) ≥60 mL/min per 1.73 m(2) were analyzed. The eGFR was calculated with the Chronic Kidney Disease Epidemiology Collaboration creatinine equation and RHF was defined as eGFR with adjusted residuals > sex-specific 95(th) percentile. RESULTS: The adjusted mean of eGFR was lower in the highest quintile of serum bicarbonate than in other quintiles, after adjusting for confounders. A lower percentile rank of serum bicarbonate was associated with higher odds of RHF. The odds ratio (OR) for RHF in the lowest quintile of serum bicarbonate was 1.39 (95 % confidence interval, 95 % CI, 1.11-1.75) compared to the highest, after adjusting for confounders. With subgroup analysis, the association was prominent in participants with a body mass index >25 kg/m(2) (OR 1.98, 95 % CI 1.32-2.95 in the lowest quintile compared to the highest), compared to those with a body mass index ≤25 kg/m(2) (OR 1.18, 95 % CI 0.89-1.56 in the lowest quintile compared to the highest). CONCLUSIONS: This study observed an association between lower serum bicarbonate and higher odds of RHF and the possible differential effect of obesity in this association. It is necessary to confirm the association between lower serum bicarbonate and RHF and its causality.

Renal hyperfiltration as a novel marker of all-cause mortality.

Although renal hyperfiltration (RHF) or an abnormal increase in GFR has been associated with many lifestyles and clinical conditions, including diabetes, its clinical consequence is not clear. RHF is frequently considered to be the result of overestimating true GFR in subjects with muscle wasting. To evaluate the association between RHF and mortality, 43,503 adult Koreans who underwent voluntary health screening at Seoul National University Hospital between March of 1995 and May of 2006 with baseline GFR≥60 ml/min per 1.73 m(2) were followed up for mortality until December 31, 2012. GFR was estimated with the Chronic Kidney Disease Epidemiology Collaboration creatinine equation, and RHF was defined as GFR>95th percentile after adjustment for age, sex, muscle mass, and history of diabetes and/or hypertension medication. Muscle mass was measured with bioimpedance analysis at baseline. During the median follow-up of 12.4 years, 1743 deaths occurred. The odds ratio of RHF in participants with the highest quartile of muscle mass was 1.31 (95% confidence interval [95% CI], 1.11 to 1.54) compared with the lowest quartile after adjusting for confounding factors, including body mass index. The hazard ratio of all-cause mortality for RHF was 1.37 (95% CI, 1.11 to 1.70) by Cox proportional hazards model with adjustment for known risk factors, including smoking. These data suggest RHF may be associated with increased all-cause mortality in an apparently healthy population. The possibility of RHF as a novel marker of all-cause mortality should be confirmed.

Quality controlled YouTube content intervention for enhancing health literacy and health behavioural intention: A randomized controlled study.

Background: Individuals increasingly turn to the Internet for health information, with YouTube being a prominent source. However, the quality and reliability of the health information vary widely, potentially affecting health literacy and behavioural intentions. Methods: To analyse the impact of health information quality on health literacy and behavioural intention, we conducted a randomized controlled trial using a quality-controlled YouTube intervention. Health information quality on YouTube was evaluated using the Global Quality Score and DISCERN. We randomly allocated (1 : 1) to the intervention group to watch the highest quality-evaluated content and to the control group to watch the lowest quality-evaluated content. Health literacy and health behavioural intention were assessed before and after watching YouTube. The trial was set for two different topics: interpreting laboratory test results from health check-up and information about inflammatory bowel disease (IBD). Results: From 8 April 2022 to 15 April 2022, 505 participants were randomly assigned to watch either high-quality content (intervention group, n = 255) or low-quality content (control group, n = 250). Health literacy significantly improved in the intervention group (28.1 before and 31.8 after; p < 0.01 for health check-up; 28.3 before and 31.3 after; p < 0.01 for IBD). Health behavioural intention significantly improved in the intervention group (3.5 before and 4.1 after; p < 0.01 for health check-up; 3.6 before and 4.0 after; p < 0.01 for IBD). Control groups had no such effect. Conclusion: High-quality health information can enhance health literacy and behavioural intention in both healthy individuals and those with specific conditions like IBD. It stresses the significance of ensuring reliable health information online and calls for future efforts to curate and provide access to high-quality health content.

Association of midlife body-weight variability and cycles with earlier dementia onset: a nationwide cohort study.

BACKGROUND: Given the rising awareness of health-related lifestyle modifications, the impact of changes in body weight (BW) on cognitive function and dementia generates significant concern. This study aimed to investigate the association between BW changes and dementia in a middle-aged Korean population. METHODS: A retrospective, population-based longitudinal study was conducted utilizing data from the National Health Insurance Service (NHIS) database. Participants aged 40 years or older in 2011 who underwent at least five health checkups between 2002 and 2011 were followed-up for dementia until 2020. A total of 3,635,988 dementia-free Korean aged < 65 at baseline were examined. We analyzed the association between BW variability independent of the mean (VIM) with BW cycle, defined as either an upward or a downward direction of BW, and the risk of incident dementia. RESULTS: The results showed an increased risk of dementia in the highest quartile of VIM quartile (hazard ratio [HR] 1.52, 95% confidence interval [CI] 1.47-1.58) compared to the lowest quartile of VIM. Additionally, the results showed an even higher increased risk of dementia in the highest BW cycle (≥ 2 cycles of 10% BW = HR 2.00, 95% CI 1.74-1.29). Notably, the combined concept of VIM with BW cycle showed an even higher dementia risk (highest quartile of VIM with ≥ 2 cycles of 10% BW = HR 2.37, 95% CI 2.05-2.74) compared to the baseline group (lowest quartile of VIM with < 3% BW cycle). CONCLUSIONS: The present study highlights the importance of considering BW changes with BW variability along with the BW cycle to assess dementia risk in detail, providing valuable insights for preventive strategies.

Trends in Childhood Glaucoma Prevalence and Incidence in South Korea, 2002-2019: A Nationwide Population-Based Study.

PRCIS: This nationwide analysis identified the prevalence and incidence of childhood glaucoma for an 18-year period. The prevalence and incidence of primary congenital glaucoma showed increasing trends. Juvenile open angle glaucoma, meanwhile, showed a decreasing tendency. PURPOSE: We aimed to determine the trends in the prevalence and incidence of childhood glaucoma in the entire population of South Korea. PATIENTS AND METHODS: A nationwide retrospective cohort study was performed with an age-specific and sex-specific population of South Korea. The Korean National Health Insurance Service claims database for 2002 to 2019 was accessed to identify cases of ophthalmologist-confirmed primary childhood glaucoma [ie, primary congenital glaucoma (PCG) and juvenile open angle glaucoma (JOAG)]. Incidence for PCG was estimated for a same-birth-year population, while that for JOAG was estimated using age-specific and sex-specific population figures. To verify the glaucoma cases, we also analyzed the diagnostic codes as well as any information on medication prescriptions and/or ocular-surgery history. RESULTS: During the 18-year observational period, totals of 505 and 7538 patients were diagnosed with PCG and JOAG, respectively. The mean prevalences of PCG and JOAG were 3.96±0.72 and 14.17±5.18, respectively. The prevalence of PCG showed an overall increasing trend during the study period, but the pattern was not significant ( ß =0.049, P =0.143); that of JOAG, meanwhile, showed a significant decreasing tendency ( ß =-0.713, P =0.001). PCG prevalence showed no difference between urban and rural areas, but JOAG showed a higher prevalence in rural areas ( P <0.001). As for mean incidence, the rates for PCG and JOAG were 1.54±0.49 and 5.02±1.95 (per 100,000 person-years), respectively, and were higher in males ( P <0.001 and P =0.013). CONCLUSION: This study identified childhood glaucoma prevalence and incidence in a general population of East Asian ethnicity. This data could help to promote a better understanding of the typical epidemiological features and clinical courses of childhood glaucoma patients.

Deep learning-based long-term risk evaluation of incident type 2 diabetes using electrocardiogram in a non-diabetic population: a retrospective, multicentre study.

Background: Diabetes is a major public health concern. We aimed to evaluate the long-term risk of incident type 2 diabetes in a non-diabetic population using a deep learning model (DLM) detecting prevalent type 2 diabetes using electrocardiogram (ECG). Methods: In this retrospective study, participants who underwent health checkups at two tertiary hospitals in Seoul, South Korea, between Jan 1, 2001 and Dec 31, 2022 were included. Type 2 diabetes was defined as glucose ≥126 mg/dL or glycated haemoglobin (HbA1c) ≥ 6.5%. For survival analysis on incident type 2 diabetes, we introduced an additional variable, diabetic ECG, which is determined by the DLM trained on ECG and corresponding prevalent diabetes. It was assumed that non-diabetic individuals with diabetic ECG had a higher risk of incident type 2 diabetes than those with non-diabetic ECG. The one-dimensional ResNet-based model was adopted for the DLM, and the Guided Grad-CAM was used to localise important regions of ECG. We divided the non-diabetic group into the diabetic ECG group (false positive) and the non-diabetic ECG (true negative) group according to the DLM decision, and performed a Cox proportional hazard model, considering the occurrence of type 2 diabetes more than six months after the visit. Findings: 190,581 individuals were included in the study with a median follow-up period of 11.84 years. The areas under the receiver operating characteristic curve for prevalent type 2 diabetes detection were 0.816 (0.807-0.825) and 0.762 (0.754-0.770) for the internal and external validations, respectively. The model primarily focused on the QRS duration and, occasionally, P or T waves. The diabetic ECG group exhibited an increased risk of incident type 2 diabetes compared with the non-diabetic ECG group, with hazard ratios of 2.15 (1.82-2.53) and 1.92 (1.74-2.11) for internal and external validation, respectively. Interpretation: In the non-diabetic group, those whose ECG was classified as diabetes by the DLM were at a higher risk of incident type 2 diabetes than those whose ECG was not. Additional clinical research on the relationship between the phenotype of ECG and diabetes to support the results and further investigation with tracked data and various ECG recording systems are suggested for future works. Funding: National Research Foundation of Korea.

Prevalence of Inflammatory Bowel Disease Unclassified, as Estimated Using the Revised Porto Criteria, among Korean Pediatric Patients with Inflammatory Bowel Disease.

Purpose: Few studies have reported the prevalence of inflammatory bowel disease unclassified (IBDU) among Korean pediatric IBD (PIBD) population. To address this gap, we used two tertiary centers and nationwide population-based healthcare administrative data to estimate the prevalence of Korean pediatric IBDU at the time of diagnosis. Methods: We identified 136 patients aged 2-17 years with newly diagnosed IBD (94 Crohn's disease [CD] and 42 ulcerative colitis [UC]) from two tertiary centers in Korea between 2005 and 2017. We reclassified these 136 patients using the revised Porto criteria. To estimate the population-based prevalence, we analyzed Korean administrative healthcare data between 2005 and 2016, which revealed 3,650 IBD patients, including 2,538 CD and 1,112 UC. By extrapolating the reclassified results to a population-based dataset, we estimated the prevalence of PIBD subtypes. Results: Among the 94 CD, the original diagnosis remained unchanged in 93 (98.9%), while the diagnosis of one (1.1%) patient was changed to IBDU. Among the 42 UC, the original diagnosis remained unchanged in 13 (31.0%), while the diagnoses in 11 (26.2%), 17 (40.5%), and one (2.4%) patient changed to atypical UC, IBDU, and CD, respectively. The estimated prevalences of CD, UC, atypical UC, and IBDU in the Korean population were 69.5%, 9.4%, 8.0%, and 13.1%, respectively. Conclusion: This study is the first in Korea to estimate the prevalence of pediatric IBDU. This prevalence (13.1%) aligns with findings from Western studies. Large-scale prospective multicenter studies on PIBDU are required to examine the clinical features and outcomes of this condition.