Journal of Korean Medical Science最新文献

This corrects the article on p. e135 in vol. 40, PMID: 40625044.

Background: Fatty liver disease is a common condition linked to metabolic syndrome, cardiovascular diseases, and liver cirrhosis, and timely, accurate diagnosis is crucial. In clinical studies, incorporating deep learning models often faces the challenge of scarce labeled data. This study investigates the effectiveness of graph-based deep learning models with attention mechanisms to predict fatty liver disease even with limited labeled data.

Methods: We utilized a dataset of 7,953 individuals, focusing on clinical variables obtained during health check-ups. Graph Neural Networks (GNNs) with attention mechanisms were assessed for predicting fatty liver disease in a semi-supervised learning setting. GNNExplainer was employed for feature importance analysis, and subgroup analysis was conducted to identify clusters with distinct risk factors.

Results: Our findings indicate that attention-based GNNs significantly outperformed conventional models in predicting fatty liver disease under semi-supervised settings, with statistically significant improvements in area under the curves (AUCs) (all P < 0.05) compared to logistic regression across most labeling scenarios. With only 10 labeled samples per class, the Graph Attention Network (GAT) and Simplified Graph Transformer with Graph Attention achieved AUCs of 0.7049 ± 0.0570 and 0.7184 ± 0.0395, respectively, and achieved AUCs of 0.7893 ± 0.0171 when 100 labeled samples were used. Feature importance analysis identified HbA1c (relative importance score = 1.0), body fat amount (0.998), and glucose (0.6934) as the most influential predictors. Subgroup analysis revealed two distinct patient clusters-one characterized by metabolic risk factors and the other by demographic and lifestyle factors-emphasizing the potential for individualized risk stratification in fatty liver disease.

Conclusion: Attention-based GNNs demonstrated strong predictive performance for fatty liver disease using a small number of labeled samples. This methodological approach illustrates how graph-based learning can leverage relational structures in routine clinical data to support data-efficient, individualized risk assessment in label-constrained settings.

Background: Highly urgent adult living donor liver transplantation (HU-LDLT) is essential for patients with acute liver failure (ALF), acute-on-chronic liver failure (ACLF), and severe cirrhosis who are in life-threatening situations. The results of adult highly urgent ABO-incompatible (ABOi) living donor liver transplantation (LDLT) remain ambiguous when there is insufficient time to await a compatible organ. This study aimed to compare the results of adult ABOi HU-LDLT with those of adult ABO-compatible (ABOc) HU-LDLT utilizing data from the Korean Network for Organ Sharing (KONOS).

Methods: We conducted a retrospective study using KONOS data from 363 consecutive adult HU-LDLT patients between 2017 and 2021 in Korea.

Results: The incidence of ABOc-LDLTs and ABOi-LDLTs was 90.6% (n = 329) and 9.4% (n = 34), respectively. Hepatitis B virus infection and alcoholism are the main etiologies of adult HU-LDLT. The median waiting time was 1 day (range, 0-36 days) for ABOc LDLT patients and 3 days (range, 0-28 days) for ABOi LDLT patients. None of the patients developed antibody-mediated rejection during follow-up. The incidence of graft failure was 17.6% in ABOi LDLT patients and 7.6% in ABOc LDLT patients. However, the overall survival and graft survival rates in the ABOi LDLT patients were not different from those in the ABOc LDLT patients. ABOi LDLT was not associated with graft failure or death in multivariable analysis.

Conclusion: The present study supports ABOi-LDLTs as a feasible and safe treatment for highly urgent patients.

Background: Seroprevalence studies monitor changes in the immune status of populations. During the coronavirus disease 2019 (COVID-19) pandemic, public health measures such as social distancing and mask-wearing led to significant changes in the epidemiology and transmission patterns of many infectious diseases. In the post-pandemic period, outbreaks of various infectious diseases have been reported globally. In this context, this study evaluated changes in varicella zoster virus (VZV) seroepidemiology in the Republic of Korea after the COVID-19 pandemic.

Methods: Residual serum samples were collected from specimens referred to Seoul Clinical Laboratories in 2023. Serum samples were collected after anonymization, with all identifiable information excluded except for date, age, sex, and region. Samples were collected evenly by age group in 5-year intervals. The evaluation of anti-VZV IgG antibodies was performed using a chemiluminescence immunoassay analyzer (Liaison^®XL; Diasorin).

Results: The study population comprised 995 participants including 581 males and 414 females. The overall VZV IgG seropositivity rate was 82.9%. The seropositivity rates of VZV IgG antibodies were 88.1% in males and 75.6% in females. VZV IgG seroprevalence showed variations across different age groups. VZV IgG seropositivity was 80% in the 1-4 year age group, declined to 53% in the 5-9 year age group, and reached its lowest at 43% in 10-14 year-olds. Seropositivity rates steadily increased from 71% in the 15-19 year-olds, to 85% in the 25-29 year group, and exceeded 95% in subjects ≥ 40 year of age.

Conclusion: Seropositivity rates in children (5-19 year) and adults (20-39 year) were lower than in previous studies in the Republic of Korea, likely due to reduced community exposure to varicella during the pandemic. These findings highlight a potential need to reinforce infection control measures and vaccination policies to address vulnerable age groups.

Background: Kidney damage can result not only from the overproduction of endogenous oxalate but also from excessive dietary intake. This study investigated whether oxalate-rich Chaga mushroom induces kidney injury.

Methods: Wistar rats were allocated to three groups based on dosese extrapolated from a previously reported clinical case. The standard-dose group received Chaga mushroom powder at 1,281.6 mg/kg body weight (equivalent to oxalate at 183 mg/kg body weight), and the high-dose (HD) group at 3,844.8 mg/kg body weight (equivalent to oxalate at 549 mg/kg body weight); the control group received no supplements. The study assessed chronic kidney injury by evaluating renal function, histopathology, oxidative stress, and apoptosis via immunohistochemistry and immunoblot assay.

Results: The final body weight of the HD group was significantly lower than that of the other groups (P = 0.011), and urinary protein excretion in the HD group was significantly higher than in the other groups (P = 0.001). Histopathologic examination revealed oxalate crystal deposition and tubular injury in the HD group. Oxidative stress markers, including 8-hydroxy-2'-deoxyguanosine levels in serum, urine, and kidney tissue from the HD group, were significantly elevated compared to other groups (P < 0.05). CD68/SR-D1 antibody levels in the HD group were significantly increased (P < 0.050). Terminal deoxynucleotidyl transferase deoxyuridine triphosphate nick end labeling and Bax positive cells in the HD group were higher than in other groups (P < 0.05), while Bcl positive cells were fewer (P < 0.05).

Conclusion: High-dose consumption of Chaga mushrooms may cause kidney damage due to its high oxalate content.

Background: Hand grip strength (HGS) has been proposed as a potential clinical marker for cognitive decline. However, its association with domain-specific cognitive changes and underlying amyloid pathology remains unclear.

Methods: This longitudinal study included 107 older adults with subjective cognitive decline (SCD) who completed a 24-month follow-up. Participants were categorized based on the presence of HGS weakness using Asian Working Group of Sarcopenia criteria. Logistic regression analyses were performed to examine the association between baseline HGS and cognitive decline across multiple domains, adjusting for relevant covariates. Repeated measures analysis of variance evaluated longitudinal changes in neuropsychological performance.

Results: Participants with HGS weakness had significantly higher amyloid positron emission tomography (PET) positivity (47.1% vs. 18.9%, P = 0.012). HGS weakness was associated with poorer baseline performance and greater decline in visuospatial and executive function over 24 months. Baseline HGS weakness predicted decline in visuospatial function (odds ratio [OR], 3.517; 95% confidence interval [CI], 1.072-11.535) and verbal memory (OR, 3.503; 95% CI, 1.046-11.729), but these associations lost significance after adjusting for amyloid positivity.

Conclusion: HGS weakness is associated with cognitive decline, particularly in visuospatial and executive domains, and may serve as an early indicator of amyloid-related neurodegeneration in older adults with SCD. HGS assessment could be a practical clinical tool for identifying individuals at risk, especially when amyloid PET is not available.

Background: The prevalence of diabetes mellitus (DM) and diabetic retinopathy (DR) is expected to increase globally; however, there have been no reports predicting DR prevalence in Korea. This study aimed to estimate the nationwide trend of the prevalence of DM and DR in Korea.

Methods: Diabetic patients aged over 40 were included from the Korean National Health and Nutrition Examination Survey database (2008-2020). The prevalence of DM and DR during the study period was calculated. Four interpolation methods were applied to address the missing DR data (2013-2016), and the method with the smallest error was selected for projection. Holt's linear trend model was used to predict DM and DR prevalence through 2040. We further projected the future number of DR patients in Korea using population projections.

Results: We included 8,007 diabetic patients aged 40 years or older and 4,540 individuals with fundus photography results. The prevalence of DM increased from 13.2% in 2008 to 17.3% (4.98 million) in 2020, while the prevalence of DR rose from 16.6% in 2008 to 24.6% (1.23 million) in 2020. The projected prevalence of DM is estimated to reach 19.7% in 2030 and 23.0% in 2040, while DR prevalence is expected to rise to 30.8% (1.97 million) in 2030 and 38.8% (3.15 million) in 2040.

Conclusion: The prevalence of DM and DR, as well as the number of patients, have increased over 13 years, and this trend is projected to continue through 2040 in Korea. Therefore, it is essential to establish effective healthcare strategies to address this trend.

Background: Despite significant morbidity and mortality, obstructive sleep apnea (OSA) remains underdiagnosed. In 2018, the South Korean National Health Insurance (NHI) expanded its coverage for polysomnography (PSG) to enhance diagnostic access for suspected OSA. This study evaluated the influence of expanded NHI coverage on PSG utilization, patient demographics, and OSA diagnosis rates in a single tertiary center.

Methods: A retrospective analysis was conducted on 1,821 adult patients who underwent in-laboratory PSG between 2015 and 2023. Demographic and clinical data, including comorbidities and PSG parameters, were collected before (pre-NHI, n = 477) and after (post-NHI, n = 1,344) coverage implementation. Patient characteristics, OSA diagnosis rates, and PSG parameters were compared between the two periods.

Results: Post-NHI, annual PSG utilization increased by 1.8-fold compared to the pre-NHI period (2015-mid-2018), with OSA diagnosis rates rising by 9.3%, reaching 84.8%, including a 7.7% increase in severe OSA cases. The mean patient age increased by 4.3 years, with a significant increase in patients aged ≥ 60 years (33.1%) and females (22.1%). Despite a stable body mass index, OSA severity metrics and comorbidities were higher post-NHI. Referral sources expanded beyond otorhinolaryngology and comprised 31.5% of the total requests post-NHI.

Conclusion: Expanded NHI coverage positively affected OSA detection and patient management, providing valuable insights into the potential for policy-driven improvements in the management of sleep disorders. The expanded PSG coverage effectively improved OSA detection in the underserved and high-risk populations.

Background: This study aimed to assess the sex-specific effects of alcohol consumption on serum uric acid (SUA) levels regarding alcohol amount, beverage types and drinking habits in Koreans.

Methods: We evaluated 17,011 adults who underwent health examinations at Samsung Medical Center between January 2011 and June 2016. Alcohol intake was measured in a standard drink unit (SDU), contained 8 g of ethanol. Individuals were categorized into six groups according to their alcohol intake from none to heavy drinkers. A dominant beverage was defined if it accounted for more than 75% of total alcohol intake. Multivariable linear regression models were used to investigate alcohol effect on SUA levels.

Results: The mean age was 51.67 ± 7.10 years, with 53.4% of men. Higher total alcohol intake was associated with higher SUA levels (P for trend < 0.001). This dose-dependent association was observed for all beverage types. Among men, heavy drinking showed a larger effect on SUA levels in the beer-dominant group (0.88 [95% confidence interval {CI}, 0.36-1.41] mg/dL per SDU) than in other dominant groups. Little soju intake in men and light beer intake in women were associated with elevated SUA levels (0.12 [95% CI, 0.03-0.20] and 0.34 [95% CI, 0.18-0.50] mg/dL per SDU, respectively). The increment in SUA levels was higher in subjects with a body mass index (BMI) < 25 kg/m² than in those with BMI ≥ 25 kg/m² across most beverage types. The effect of soju intake on SUA levels was more than twice as high in women compared to men (0.07 [95% CI, 0.02-0.11] versus 0.03 [95% CI, 0.02-0.04] mg/dL per SDU) in the BMI < 25 kg/m² group.

Conclusion: Dose-dependent associations between SUA levels and alcohol consumption were consistently observed across different beverage types, drinking habits, sex, and BMI, even at low dose alcohol intake. These findings may aid healthcare providers to offer personalized guidance on alcohol consumption for individuals with hyperuricemia.

Background: Hip fractures in older adults are associated with high rates of mortality, functional decline, and secondary fractures. Although Fracture Liaison Services (FLSs) have shown clinical benefits in many countries, a comprehensive, coordinator-based FLS model has not been widely implemented in Korea. This study aimed to evaluate the clinical impact of a newly introduced coordinator-based FLS in elderly hip fracture patients.

Methods: This prospective cohort study included patients aged 65 and older who were admitted with a hip fracture to a single tertiary hospital between June 2022 and February 2024. Patients were divided into two groups: those who received FLS after July 2023 (n = 105) and those who did not (n = 168). Clinical data were collected during hospitalization and at 6 weeks, 3 months, 6 months, and 1 year postoperatively. Variables included time to surgery, length of hospital stay, mortality, refracture rate, osteoporosis treatment rates, functional and nutritional outcomes, and patient satisfaction.

Results: The FLS group had significantly shorter time to surgery (2.5 ± 2.3 vs. 4.4 ± 5.5 days, P < 0.001) and hospital stay (20.0 ± 11.3 vs. 24.7 ± 18.0 days, P = 0.010). In-hospital mortality (1.0% vs. 4.2%), 6-month (4.8% vs. 6.5%), and 1-year mortality (8.6% vs. 12.5%) were all lower in the FLS group (P < 0.05). Refracture rates at 6 months (2.8% vs. 7.1%) and 1 year (5.7% vs. 10.7%) were also significantly lower in the FLS group (P < 0.05). Osteoporosis medication prescription (68.6% vs. 48.8%) and calcium/vitamin D supplementation (63.8% vs. 15.9%) were significantly higher in the FLS group (P < 0.001). Although no significant differences were observed in functional recovery, the FLS group showed a significant increase in serum albumin over 1 year (P = 0.022). Patient satisfaction exceeded 90% at all follow-up intervals.

Conclusion: The coordinator-based FLS service reduced the length of hospital stay and time from admission to surgery in elderly hip fracture patients, while also lowering the risk of postoperative mortality and refracture. It increased the prescription rate of osteoporosis medications and can improve patients' nutritional status. However, further research is needed to assess functional improvement.