BACKGROUND: Hypertension is a serious chronic disease that can significantly lead to various cardiovascular diseases, affecting vital organs such as the heart, brain, and kidneys. Our goal is to predict the risk of new onset hypertension using machine learning algorithms and identify the characteristics of patients with new onset hypertension. METHODS: We analyzed data from the 2011 China Health and Nutrition Survey cohort of individuals who were not hypertensive at baseline and had follow-up results available for prediction by 2015. We tested and evaluated the performance of four traditional machine learning algorithms commonly used in epidemiological studies: Logistic Regression, Support Vector Machine, XGBoost, LightGBM, and two deep learning algorithms: TabNet and AMFormer model. We modeled using 16 and 29 features, respectively. SHAP values were applied to select key features associated with new onset hypertension. RESULTS: A total of 4,982 participants were included in the analysis, of whom 1,017 developed hypertension during the 4-year follow-up. Among the 16-feature models, Logistic Regression had the highest AUC of 0.784(0.775∼0.806). In the 29-feature prediction models, AMFormer performed the best with an AUC of 0.802(0.795∼0.820), and also scored the highest in MCC (0.417, 95%CI: 0.400∼0.434) and F1 (0.503, 95%CI: 0.484∼0.505) metrics, demonstrating superior overall performance compared to the other models. Additionally, key features selected based on the AMFormer, such as age, province, waist circumference, urban or rural location, education level, employment status, weight, WHR, and BMI, played significant roles. CONCLUSION We used the AMFormer model for the first time in predicting new onset hypertension and achieved the best results among the six algorithms tested. Key features associated with new onset hypertension can be determined through this algorithm. The practice of machine learning algorithms can further enhance the predictive efficacy of diseases and identify risk factors for diseases.
Humans ; China/epidemiology* ; Hypertension/diagnosis* ; Machine Learning ; Male ; Female ; Middle Aged ; Adult ; Nutrition Surveys ; Algorithms ; Aged ; Risk Factors

Objective: To explore the influencing factors of pregnancy-induced hypertensive disorders in pregnancy (HDP) with organ or system impairment in pregnant women, and to analyze and compare the differences of HDP subtypes in different regions of China. Methods: A total of 27 680 pregnant women with HDP with complete data from 161 hospitals in 24 provinces, autonomous regions and municipalities were retrospectively collected from January 1, 2018 to December 31, 2018. According to their clinical manifestations, they were divided into hypertension group [a total of 10 308 cases, including 8 250 cases of gestational hypertension (GH), 2 058 cases of chronic hypertension during pregnancy] and hypertension with organ or system impairment group [17 372 cases, including 14 590 cases of pre-eclampsia (PE), 137 cases of eclampsia, 2 645 cases of chronic hypertension with PE]. The subtype distribution of HDP in East China (6 136 cases), North China (4 821 cases), Central China (3 502 cases), South China (8 371 cases), Northeast China (1 456 cases), Southwest China (2 158 cases) and Northwest China (1 236 cases) were analyzed. By comparing the differences of HDP subtypes and related risk factors in different regions, regional analysis of the risk factors of HDP pregnant women with organ or system impairment was conducted. Results: (1) The proportions of HDP pregnant women with organ or system impairment in Northeast China (79.05%, 1 151/1 456), Central China (68.42%, 2 396/3 502) and Northwest China (69.34%, 857/1 236) were higher than the national average (62.76%, 17 372/27 680); the proportions in North China (59.18%, 2 853/4 821), East China (60.85%, 3 734/6 136) and South China (59.56%, 4 986/8 371) were lower than the national average, and the differences were statistically significant (all P<0.05). (2) Univariate analysis showed that the proportions of primiparas, non-Han, non-urban household registration, irregular prenatal examination and PE history in the hypertension with organ or system impairment group were higher than those in the hypertension group, and the differences were statistically significant (all P<0.05). Multivariate logistic regression analysis showed that primiparas, non-Han, non-urban household registration, irregular prenatal examination and PE history were independent risk factors for HDP pregnant women with organ or system impairment (all P<0.05). (3) Primipara: the rates of primipara in Northeast China, North China and Southwest China were higher than the national average level, while those in South China, Central China and Northwest China were lower than the national average level. Non-Han nationality: the rates of non-Han nationality in Northeast China, North China and Northwest China were higher than the national average, while those in East China, South China and Central China were lower than the national average. Non-urban household registration: the rates of non-urban household registration in Northeast China, North China, and Southwest China were lower than the national average, while those in East China, Central China were higher than the national average. Irregular prenatal examination: the rates of irregular prenatal examination in North China, South China and Southwest regions were lower than the national average level, while those in Northeast China, Central China and Northwest China were higher than the national average level. History of PE: the incidence rates of PE in Northeast China, North China, South China and Southwest China were lower than the national average level, while those in Central China and Northwest China were higher than the national average level. Conclusions: Primiparas, non-Han, non-urban household registration, irregular prenatal examination, and PE history are risk factors for HDP pregnant women with organ or system impairment. Patients in Northeast, Central and Northwest China have more risk factors, and are more likely to be accompanied by organ or system function damage. It is important to strengthen the management of pregnant women and reduce the occurrence of HDP.
Humans ; Pregnancy ; Female ; Hypertension, Pregnancy-Induced/diagnosis* ; Retrospective Studies ; Pre-Eclampsia/epidemiology* ; Risk Factors ; Incidence

BACKGROUND: The ankle-brachial index (ABI) is a simple, noninvasive clinical test that is useful in the diagnosis of peripheral artery disease (PAD). The borderline ABI value is 0.91–0.99 and has also been reported in endothelial dysfunction and preclinical PAD. We investigated the relationship between resting heart rate as a surrogate marker of autonomic balance and borderline ABI in apparently healthy individuals.METHODS: We evaluated the association between resting heart rate and borderline ABI in 618 participants (348 male and 270 female) in a health examination program. The odds ratios for borderline ABI were calculated using a multivariable logistic regression analysis after adjusting for confounding variables across heart rate quartiles (Q1≤56, Q2=57–62, Q3=63–68, and Q4≥69 beats/min).RESULTS: The overall prevalence of borderline ABI was 13.9%. The age- and sex-adjusted resting heart rate was significantly higher in the borderline ABI group than in the control group (66.9±0.9 vs. 63.0±0.4 beats/min, P < 0.001). The odds ratios (95% confidence intervals) for the borderline ABI in each heart rate quartile were 1.00, 1.04 (0.43–2.52), 1.69 (0.73–3.93), and 3.52 (1.55–7.97) after adjusting for age, sex, smoking status, alcohol intake, regular exercise, body mass index, systolic blood pressure, fasting plasma glucose level, triglyceride level, high-density lipoprotein-cholesterol level, white blood cell count, C-reactive protein level, uric acid level, γ-glutamyltransferase level, hypertension medication, diabetes medication, and dyslipidemia medication.CONCLUSION: These findings indicate that a higher resting heart rate is independently associated with borderline ABI.
Ankle Brachial Index ; Biomarkers ; Blood Glucose ; Blood Pressure ; Body Mass Index ; C-Reactive Protein ; Confounding Factors (Epidemiology) ; Diagnosis ; Dyslipidemias ; Fasting ; Heart Rate ; Heart ; Humans ; Hypertension ; Leukocyte Count ; Logistic Models ; Male ; Odds Ratio ; Peripheral Arterial Disease ; Prevalence ; Smoke ; Smoking ; Triglycerides ; Uric Acid

BACKGROUND/AIMS: The aim of this study was to assess the association between the number of existing permanent teeth and chronic kidney disease (CKD) in a representative sample of the elderly Korean population. METHODS: A total of 2,519 subjects who participated in the Korean National Health and Nutrition Examination Survey were cross-sectionally examined. The number of existing permanent teeth was evaluated by clinical oral examination. CKD was defined based on definition and classification by Kidney Disease: Improving Global Outcomes (KDIGO) 2012 guidelines. Multivariable logistic regression analyses were performed controlling for age, gender, income, education, tooth-brushing frequency, periodontitis, state of dentition, smoking, alcohol consumption, hypertension, obesity, diabetes mellitus, and hypercholesterolemia. Subgroup analyses by age and gender were also performed. RESULTS: The number of teeth was significantly associated with CKD after controlling for all potential confounders (adjusted odds ratio [AOR], 1.67; 95% confidence interval [CI], 1.04 to 2.70 for lower number of teeth; AOR, 1.59; 95% CI, 1.14 to 2.23 for moderate number of teeth). In the subgroup analyses, the association was highlighted in females aged 75 years over (AOR, 2.55; 95% CI, 1.05 to 6.20 for lower number of teeth; AOR, 1.95; 95% CI, 1.01 to 3.80 for moderate number of teeth). CONCLUSIONS: Our findings suggest that the number of existing permanent teeth may be associated with CKD among Korean elderly.
Aged* ; Alcohol Drinking ; Classification ; Dentition ; Diabetes Mellitus ; Diagnosis, Oral ; Education ; Epidemiology ; Female ; Humans ; Hypercholesterolemia ; Hypertension ; Kidney Diseases ; Logistic Models ; Nutrition Surveys ; Obesity ; Odds Ratio ; Periodontitis ; Renal Insufficiency, Chronic* ; Risk Factors ; Smoke ; Smoking ; Tooth*

OBJECTIVES: To identify the most important demographic risk factors for a diagnosis of type 2 diabetes mellitus (T2DM) using a neural network model. METHODS: This study was conducted on a sample of 234 individuals, in whom T2DM was diagnosed using hemoglobin A1c levels. A multilayer perceptron artificial neural network was used to identify demographic risk factors for T2DM and their importance. The DeLong method was used to compare the models by fitting in sequential steps. RESULTS: Variables found to be significant at a level of p < 0.2 in a univariate logistic regression analysis (age, hypertension, waist circumference, body mass index [BMI], sedentary lifestyle, smoking, vegetable consumption, family history of T2DM, stress, walking, fruit consumption, and sex) were entered into the model. After 7 stages of neural network modeling, only waist circumference (100.0%), age (78.5%), BMI (78.2%), hypertension (69.4%), stress (54.2%), smoking (49.3%), and a family history of T2DM (37.2%) were identified as predictors of the diagnosis of T2DM. CONCLUSIONS: In this study, waist circumference and age were the most important predictors of T2DM. Due to the sensitivity, specificity, and accuracy of the final model, it is suggested that these variables should be used for T2DM risk assessment in screening tests.
Body Mass Index ; Diabetes Mellitus, Type 2* ; Diagnosis ; Epidemiology ; Fruit ; Humans ; Hypertension ; Iran ; Logistic Models ; Mass Screening ; Methods ; Models, Statistical ; Neural Networks (Computer)* ; Risk Assessment ; Risk Factors* ; Sedentary Lifestyle ; Sensitivity and Specificity ; Smoke ; Smoking ; Vegetables ; Waist Circumference ; Walking

OBJECTIVES: To identify the most important demographic risk factors for a diagnosis of type 2 diabetes mellitus (T2DM) using a neural network model.METHODS: This study was conducted on a sample of 234 individuals, in whom T2DM was diagnosed using hemoglobin A1c levels. A multilayer perceptron artificial neural network was used to identify demographic risk factors for T2DM and their importance. The DeLong method was used to compare the models by fitting in sequential steps.RESULTS: Variables found to be significant at a level of p < 0.2 in a univariate logistic regression analysis (age, hypertension, waist circumference, body mass index [BMI], sedentary lifestyle, smoking, vegetable consumption, family history of T2DM, stress, walking, fruit consumption, and sex) were entered into the model. After 7 stages of neural network modeling, only waist circumference (100.0%), age (78.5%), BMI (78.2%), hypertension (69.4%), stress (54.2%), smoking (49.3%), and a family history of T2DM (37.2%) were identified as predictors of the diagnosis of T2DM.CONCLUSIONS: In this study, waist circumference and age were the most important predictors of T2DM. Due to the sensitivity, specificity, and accuracy of the final model, it is suggested that these variables should be used for T2DM risk assessment in screening tests.
Body Mass Index ; Diabetes Mellitus, Type 2 ; Diagnosis ; Epidemiology ; Fruit ; Humans ; Hypertension ; Iran ; Logistic Models ; Mass Screening ; Methods ; Models, Statistical ; Neural Networks (Computer) ; Risk Assessment ; Risk Factors ; Sedentary Lifestyle ; Sensitivity and Specificity ; Smoke ; Smoking ; Vegetables ; Waist Circumference ; Walking

INTRODUCTION: This study aimed to determine the 5-year incidence of albuminuria among Asian persons with newly diagnosed type 2 diabetes mellitus (DM), and to identify the risk factors at diagnosis for progression to albuminuria. MATERIALS AND METHODS: A retrospective 5-year closed cohort study was conducted among 1016 persons aged ≥18 years old who were diagnosed with type 2 DM between 1 January 2007 and 31 December 2009 at primary care facilities in Singapore. The cumulative incidence of progression from normoalbuminuria to albuminuria-termed "progression"-was determined. The risk factors associated with progression were evaluated using multiple logistic regression analysis. RESULTS: A total of 541 (53.2%) participants were men. The mean (SD) onset age of type 2 DM was 54 (11) years. From diagnosis of type 2 DM, the 5-year cumulative incidence of progression was 17.3% and mean (SD) duration to progression was 2.88 (1.23) years. Higher onset age (OR 1.02; 95% CI, 1.00-1.04), history of hypertension (OR, 1.88; 95% CI, 1.32-2.70) and higher glycated haemoglobin (HbA1c) (OR, 1.17; 95% CI, 1.09-1.26) at diagnosis were associated with progression. In addition, being on angiotensin converting enzyme inhibitors (ACEIs) or angiotensin receptor blockers (ARBs) treatment at baseline modified the effect of hypertension on progression. CONCLUSION This study highlighted the importance of early screening and treatment of diabetes as well as prevention of hypertension, which could potentially delay the onset of microalbuminuria in persons with type 2 DM. Persons on ACEI or ARB treatment should continue to be monitored regularly for progression to albuminuria.
Adult ; Age of Onset ; Aged ; Albuminuria ; epidemiology ; Angiotensin Receptor Antagonists ; therapeutic use ; Angiotensin-Converting Enzyme Inhibitors ; therapeutic use ; Cohort Studies ; Diabetes Mellitus, Type 2 ; diagnosis ; epidemiology ; metabolism ; Disease Progression ; Female ; Glycated Hemoglobin A ; metabolism ; Humans ; Hypertension ; drug therapy ; epidemiology ; Logistic Models ; Male ; Middle Aged ; Retrospective Studies ; Risk Factors ; Singapore ; epidemiology

BACKGROUND: To date, numerous studies have investigated the epidemiology and clinical characteristics of male pattern hair loss (MPHL). OBJECTIVE: The purpose of this study was to evaluate a large number of long-term MPHL patients and to compare them to shorter term ones from other studies. METHODS: We retrospectively reviewed charts of patients diagnosed with MPHL during a 10-year period at the alopecia clinic at the Department of Dermatology of Myongji Hospital. RESULTS: Among 3,549 alopecia patients who had visited the hospital's alopecia clinic, 1,360 were patients demonstrating MPHL (38.3%). The most frequent age group was patients in their 20's (30.1%) followed by those in their 30's (28.2%), 40's (17.2%), 50's (9.9%), and teenagers (8.9%). There were 599 MPHL patients (44.0%) with a paternal familial predisposition, 182 (13.4%) with a maternal familial predisposition, and 118 (8.7%) with familial predisposition on maternal and paternal sides. MPHL was classified based on the Norwood-Hamilton classification: type 2 (20.0%), type 3v (19.4%), and type 3a (16.3%) were the most frequent. Abnormalities in total serum cholesterol were found in 15.1% patients and triglycerides in 36.1% of patients. The most common comorbidity was seborrheic dermatitis, followed by hypertension, hyperlipidemia, diabetes mellitus, and gastrointestinal diseases. CONCLUSION: This study demonstrated that the period of MPHL diagnosis and treatment becomes earlier in age and milder regarding MPHL type, which may be due to early puberty in teenagers.
Adolescent ; Alopecia ; Cholesterol ; Classification ; Comorbidity ; Dermatitis, Seborrheic ; Dermatology ; Diabetes Mellitus ; Diagnosis ; Epidemiology ; Gastrointestinal Diseases ; Hair* ; Humans ; Hyperlipidemias ; Hypertension ; Male* ; Puberty ; Retrospective Studies ; Triglycerides

BACKGROUND: Active patient participation in health care decision-making can results in better medical outcomes. This study's purpose is to investigate the association between the patient participation and the diseases often encountered in primary care. METHODS: We used the data from the Korean National Health and Nutrition Examination Survey conducted in 2015 which included 4,158 adults aged older than 19 and who had no missing data. The association between the presence of disease or the number of accompanying diseases and the active patient participation in the treatment decision was studied. Logistic regression analysis was conducted using complex sampling design in each sex. RESULTS: After adjusting for confounding variables, the relationship between active patient participation and the diagnosis of hypertension, odds ratio (OR) was 1.95 (95% confidence interval [CI], 1.25-3.04) for men and 1.83 (95% CI, 1.27-2.65) for women. In women diagnosed with diabetes, OR was 0.58 (95% CI, 0.35-0.97). Between active patient participation and increasing number of accompanying diseases have positive tendency but not statistically significant. CONCLUSIONS: This study suggests that active patient participation is related to the diagnosis of hypertension, and the number of accompanying diseases and active patient participation were not significantly associated. This is a rare study related to the active patient participation in the Korean population, that it may be helpful in establishing further relevant research and strategies to increase the patient participation rate.
Adult ; Chronic Disease* ; Compliance ; Confounding Factors (Epidemiology) ; Delivery of Health Care ; Diagnosis ; Female ; Humans ; Hypertension ; Logistic Models ; Male ; Nutrition Surveys ; Odds Ratio ; Patient Participation* ; Primary Health Care

The Korean Genome Epidemiology Study (KoGES) is a principal cohort study providing valuable evidence for the prevention of major chronic diseases such as hypertension, obesity, and diabetes in Korea. Since 2001, the Ansan-Anseong cohort is one of the representative cohorts in the KoGES and recruited about 10,000 participants from Ansan and Anseong city to undergo a comprehensive health examination biennially. About 3,000 participants in the Ansan cohort underwent abdominal computed tomography scan to detect the presence of nonalcoholic fatty liver disease (NAFLD). The prevalence of NAFLD was about 23% in this study, and it was twice as high in subjects with diabetes compared to those without. Subjects with NAFLD had early diastolic dysfunction in tissue Doppler study and showed lower vitamin D concentrations than those without. We also found that the palatin-like phospholipase domain containing 3 single nucleotide polymorphisms were significantly associated with NAFLD. Interim prospective analysis after six years showed that NAFLD was associated with worsening of metabolic risk factors and an about 2.6 higher likelihood of developing diabetes than in those without. These results present the clinical importance of the diagnosis and treatment of NAFLD to reduce the future development of diabetes and cardiovascular diseases.
Cardiovascular Diseases ; Chronic Disease ; Cohort Studies ; Diagnosis ; Epidemiology* ; Fatty Liver ; Genome* ; Gyeonggi-do ; Hypertension ; Korea ; Non-alcoholic Fatty Liver Disease* ; Obesity ; Phospholipases ; Polymorphism, Single Nucleotide ; Prevalence ; Prospective Studies ; Risk Factors ; Vitamin D