OBJECTIVE To explore the construction of the physical identification model of twenty-five people of yin and yang by combining the hand diagnosis image analysis of machine learning,in order to enrich the input of physical identification features.METHODS Based on the theory of twenty-five people of yin and yang,the hand images of 542 patients with cardiovascular disease were collected by self-developed hand diagnosis acquisition equipment.The key points of the hand were extracted by the Mediapipe palm key point algorithm,and the palm area was segmented by geometry to obtain the length and color gamut characteristics of different parts.The physician with clinical license was used to determined the patient's yin and yang twenty-five constitution according to the e-lectronic medical record,and the patient's constitution was identified by comparing Random Forest,Logistic Regression,Xgboost algo-rithm and Light GBM.RESULTS This study successfully constructed a TCM constitution identification model based on hand charac-teristics,which can identify ten basic constitution types including gold wood,gold fire,gold water,gold soil,wood fire,wood water,wood soil,water fire,water soil and fire soil.After sample equalization,the Random Forest model performed best in physical identifi-cation,with an accuracy rate of 0.69,which was significantly better than other models.CONCLUSION The constitution identifica-tion model of TCM hand diagnosis based on hand characteristics can realize the constitution identification of the five elements of TCM.However,due to the use of only hand diagnosis data of TCM,the accuracy needs to be improved.Compared with the traditional body mass scale,it is more objective,convenient,highly stable and consistent.In the future,multimodal integration with face diagno-sis and whole body inspection can be carried out to promote the accuracy,intelligence and multi-dimensional evaluation ability of TCM constitution identification technology.

Objective This article aims to investigate the association between hypertension and the risk of GSD by conducting a national multicenter study,a systematic review,and a meta-analysis.Methods The study was conducted in three stages.In the first stage,subjects were recruited for health examination in four hospitals in Chengdu,Tianjin,Beijing,and Chongqing,China,from 2015 to 2020,and the multivariate logistic regression analysis was used to investigate the association between hypertension and the risk of GSD in each center.In the second stage,Embase,PubMed,Wanfang Data,VIP,and CNKI databases were searched for related studies published up to May 2021,and a meta-analysis was conducted to further verify such association.In the third stage,the random effects model was used for pooled analysis of the results of the multicenter cross-sectional study and the findings of previous literature.Results A total of 633 948 participants were enrolled in the cross-sectional study,and the prevalence rate of GSD was 7.844%.The multivariate logistic regression analysis showed that hypertension was positively associated with the risk of GSD(P＜0.05).Subgroup analysis showed that there was no significant difference in the association between hypertension and GSD between individuals with different sexes,ages,and subtypes of GSD.A total of 80 articles were included in the systematic review and the meta-analysis,and the results showed that the risk of GSD was increased by 1.022 times for every 10 mmHg increase in diastolic pressure and 1.014 times for every 10 mmHg increase in systolic pressure.Conclusion Hypertension significantly increases the risk of GSD,and the findings of this study will provide a basis for the etiology of GSD and the identification of high-risk groups.

OBJECTIVE To explore the construction of the physical identification model of twenty-five people of yin and yang by combining the hand diagnosis image analysis of machine learning,in order to enrich the input of physical identification features.METHODS Based on the theory of twenty-five people of yin and yang,the hand images of 542 patients with cardiovascular disease were collected by self-developed hand diagnosis acquisition equipment.The key points of the hand were extracted by the Mediapipe palm key point algorithm,and the palm area was segmented by geometry to obtain the length and color gamut characteristics of different parts.The physician with clinical license was used to determined the patient's yin and yang twenty-five constitution according to the e-lectronic medical record,and the patient's constitution was identified by comparing Random Forest,Logistic Regression,Xgboost algo-rithm and Light GBM.RESULTS This study successfully constructed a TCM constitution identification model based on hand charac-teristics,which can identify ten basic constitution types including gold wood,gold fire,gold water,gold soil,wood fire,wood water,wood soil,water fire,water soil and fire soil.After sample equalization,the Random Forest model performed best in physical identifi-cation,with an accuracy rate of 0.69,which was significantly better than other models.CONCLUSION The constitution identifica-tion model of TCM hand diagnosis based on hand characteristics can realize the constitution identification of the five elements of TCM.However,due to the use of only hand diagnosis data of TCM,the accuracy needs to be improved.Compared with the traditional body mass scale,it is more objective,convenient,highly stable and consistent.In the future,multimodal integration with face diagno-sis and whole body inspection can be carried out to promote the accuracy,intelligence and multi-dimensional evaluation ability of TCM constitution identification technology.

Objective:To analyze the relationship between dietary composition of residents in endemic fluorosis areas and skeletal fluorosis.Methods:A case-control study was used to analyze the difference of dietary composition between patients with skeletal fluorosis (case group) and residents without skeletal fluorosis (control group). In August 2019, taking the drinking water-borne endemic fluorosis area in Wenshui County, Lvliang City, Shanxi Province as the survey site, a cluster sampling method was adopted to select local residents aged over 18 years old, and a questionnaire survey was conducted by face-to-face interview. The survey contents included gender, age and consumption frequency of various foods. Binary logistic regression was used to analyze the relationship between food consumption frequency and skeletal fluorosis. The diagnosis of skeletal fluorosis was made by using portable digital radiography (DR) to take X-ray films of forearm and lower leg, combining with clinical signs, and according to the Diagnostic Standard for Endemic Skeletal Fluorosis (WS/T 192-2008) to determine.Results:A total of 1 061 subjects were included in this study, including 376 in the case group and 685 in the control group. The age composition of patients in the case group (≤60, > 60 years old: 162, 214 cases) was significantly different from that in the control group (≤60, > 60 years old: 423, 261 cases, χ 2 = 34.52, P < 0.001). There was no statistically significant difference in gender ratio (χ 2 = 1.37, P = 0.251). The proportion of patients in the case group who ate meat and eggs > 1 time/week was lower than that in the control group (χ 2 = 8.06, 5.46, P < 0.05), the proportion of patients who ate milk > 1 time/week was higher than that in the control group (χ 2 = 4.01, P = 0.046), and the proportion of patients who ate seafood ≥1 time/week was lower than that in the control group (χ 2 = 4.16, P = 0.046). The results of binary logistic regression analysis showed that after adjusting for age, sex, and urinary fluoride, the frequency of eating meat, eggs or milk > 1 time/week and the frequency of eating seafood ≥1 time/week were not related to the risk of skeletal fluorosis ( P > 0.05); however, in the group ≤60 years old, the frequency of eating eggs > 1 time/week was associated with the risk of skeletal fluorosis [odds ratio ( OR) = 0.59, 95% confidence interval ( CI): 0.39, 0.88]. Conclusions:The consumption frequency of meat, milk, eggs and seafood is significantly different between the skeletal fluorosis patients and the control people. In the population ≤60 years old, consumption frequency of eggs > 1 time/week may reduce the risk of skeletal fluorosis.

Objective:Through determination of urinary arsenic metabolites in high water arsenic exposed areas of Jilin and Shanxi provinces, to explore the mode and possible influencing factors of arsenic metabolism in different populations.Methods:From October 2018 to August 2019, a cluster sampling was carried out in villages (arsenic in drinking water ≥0.05 mg/L) of some townships (towns) in Lyuliang City, Shanxi Province and Baicheng City, Jilin Province for epidemiological investigation and general health examination. The residents over 35 years old drinking water from local centralized water supply and small well water sources were selected as arsenic exposure group, and people (nearby low-arsenic water source areas) with the same diet and living habits and similar economic conditions were selected as control group. Urine samples were collected. Liquid chromatography-atomic fluorescence spectrometry(LC-AFS) technology was used to separate and detect 4 species of arsenic compounds, including trivalent inorganic arsenic (iAs Ⅲ), pentavalent inorganic arsenic (iAs Ⅴ), methylated arsine (MMA), and dimethylated arsine (DMA). Total arsenic (tAs), inorganic arsenic percentage (iAs%), MMA percentage (MMA%), DMA percentage (DMA%), primary methylation index (PMI) and the secondary methylation index (SMI) were calculated. The influencing factors of arsenic metabolism were analyzed by multiple linear regression. Results:A total of 1 415 villagers were investigated, including 1 256 in arsenic exposure group and 159 in control group. Compared with the control group, there were no significant differences in age, gender ratio and occupation distribution between arsenic exposure group and control group ( P > 0.05), but there were significant differences in smoking, drinking, body mass index (BMI) and education level distribution ( P < 0.05). The median of urinary tAs, iAs%, MMA%, DMA%, PMI and SMI in control group and arsenic exposure group were 12.86 μg/L, 15.03, 5.23, 76.35, 84.97, 93.68 and 69.68 μg/L, 10.24, 8.37, 79.31, 89.76, 90.65, respectively, the levels of urinary tAs, DMA% and PMI in arsenic exposed group were higher than those in control group, while iAs% and SMI were lower than those in control group, the differences were statistically significant ( U=- 13.87, - 4.30, - 6.64, - 6.64, - 1.99, P < 0.05). After analysis of the factors influencing urinary arsenic metabolism in the population, we found that age and BMI had an impact on iAs% ( β=- 0.08, - 0.08, P < 0.05); gender, drinking, BMI and education level were influencing factors of MMA% ( β =- 0.11, - 0.09, - 0.07, 0.08, P < 0.05); DMA% was mainly affected by age, gender, BMI and education level ( β = 0.06, 0.09, 0.10, - 0.09, P < 0.05); PMI was mainly affected by age and BMI ( β = 0.08, 0.08, P < 0.05); while SMI was affected by gender, drinking, BMI and education level ( β=0.09, 0.08, 0.08, - 0.09, P < 0.05). Conclusions:The urinary arsenic metabolism models of different arsenic exposed groups are different. Age, gender, smoking, drinking, BMI and education level may be influencing factors of different arsenic metabolism models.

Objective:To investigate the association of single nucleotide polymorphism at the estrogen receptor 1(ESR1) gene rs1801132 with the risk of brick-tea type skeletal fluorosis.Methods:The typical brick-tea type fluorosis areas in Qinghai, Xinjiang, and Inner Mongolia were selected as the survey sites for a cross-sectional study. An epidemiological questionnaire was conducted by the staffs on the sites for participants older than 16 years, and physical examination and X-ray diagnosis were performed. Brick tea, blood, and urine samples were collected at the same time. The diagnosis of skeletal fluorosis through X-ray was based on the "Diagnostic Criteria for Endemic Skeletal Fluorosis" (WS/T 192-2008); The determination of tea's fluoride and urinary fluoride was performed by fluoride ion-selective electrode method; gene sequencing analysis of rs1801132 locus of ESR1 gene was done by Sequenom MassARRAY flight mass spectrometry system.Results:A total of 994 patients were included in this study. The total prevalence of skeletal fluorosis was 23.9% (238/994). The prevalence of skeletal fluorosis in Tibetans(39.9%, 123/308) was higher than those of Mongolian and Han nationality [22.2% (58/261), 13.4% (57/425), χ 2=20.435, 67.811, P < 0.05]. Based on binary logistic analysis, the daily tea fluoride intake ≤ 3.5 mg, urinary fluoride content ≤1.6 mg/L, and age ≤45 years were used as the reference groups, and then, when the daily tea fluoride intake > 7.0 mg ( OR=2.865, 95% CI: 1.923-4.268), urinary fluoride content > 1.6-3.2 mg/L ( OR=2.368, 95% CI: 1.686-3.326) and > 3.2 mg/L ( OR=3.559, 95% CI: 2.401-5.276), the age > 45-65 years old ( OR=2.361, 95% CI: 1.603-3.477) and > 65 years old ( OR=4.556, 95% CI: 2.845-7.296), the risk of fluorosis was higher than that of the reference group, respectively. When the daily tea fluoride intake was > 3.5-7.0 mg and the level of urinary fluoride was > 1.6-3.2 mg/L, G allele had a protective effect on skeletal fluorosis in Mongolian population (adjusted OR=0.207, 95% CI: 0.044-0.974); when the daily tea fluoride intake was > 3.5-7.0 mg, gender was male group, G allele had a protective effect on skeletal fluorosis in Han population (adjusted OR=0.315, 95% CI: 0.112-0.887). Conclusion:The single nucleotide polymorphism of the rs1801132 locus at the ESR1 gene may be associated with the risk of susceptibility to brick-tea type skeletal fluorosis in Mongolian and Han nationality.

Objective: To analyze the relationship between family behaviors and overweight/obesity in primary and junior school students aged 6-14 years in Xuzhou, and to provide a reference for a targeted measure to prevent and control overweight and obesity. Methods: Using multistage stratified cluster random sampling, a total of 6 220 students aged 6-14 years old from 10 primary schools and 10 junior schools were investigated by a self-designed questionnaire. Chi-square and multivariate Logistic regression models were used to explore the relationship between family behaviors and overweight/obesity in primary and junior school students. Results: The rate of overweight/obesity in primary and junior boys was higher than that in primary and junior girls. The rate of overweight/obesity in urban students was higher than that of rural students(P<0.05). The Chi-square analysis showed that overweight of parents, irregular breakfast, eating fast food, eating sweets, drinking sweetened beverage, long screen time and short sleep duration were risk family behavior factors of overweight/obesity in primary and junior boy students(P<0.05). The risk family behavior factors of overweight/obesity in primary and junior girl students were overweight of parents, irregular breakfast, eating fast food and eating sweets(P<0.05). The risk family behavior factors of overweight/obesity, such as drinking sweetened beverage and short sleep duration, were also related to primary girls(P<0.05), and long screen time was related to junior girls(P<0.05). The multivariate Logistic regression showed that such family behavior factors as irregular breakfast(OR-boy=1.58, OR-girl=1.74), eating fast food(OR-boy=1.37, OR-girl=1.11), eating sweets(OR-boy=1.85, OR-girl=1.52), drinking sweetened beverage(OR-boy=1.64, OR-girl=1.33) and short sleep duration(OR-boy=1.56, OR-girl=1.69) were positively correlated with the prevalence of overweight/obesity in primary students. Long screen time was also correlated to overweight/obesity primary boy students(OR=1.18). Family behavior factors for child overweight and obesity induded overweight of parents(OR-boy=1.29, OR-girl=1.23) and eating sweets(OR-boy=1.44, OR-girl=1.51). Irregular breakfast(OR=1.51), eating fast food(OR=1.22), drinking sweetened beverage (OR=1.75) and long visual screen time (OR=1.15) were also positively correlated with the prevalence of overweight/obesity in junior boy students. Conclusion Family behavior factors were positively correlated with the prevalence of overweight/obesity in primary and junior students. The influence of family behavior factors were different between primary and junior students. Behavioral interventions based on family should be adopted to prevent and control the overweight/obesity of children.

Objective To analyze the relationship between sleep duration, screen viewing time, and the prevalence of overweight/obesity among primary school students in Xuzhou. Methods Using a cluster sampling method, a total of 3 228 students (including 1 679 boys and 1 549 girls with an average age of 10.78±0.69 years) from grade one to six from 10 primary schools in Xuzhou underwent interview using a self?designed questionnaire containing basic characteristics, sleep duration, and screen viewing time. Data on height and weight were also collected. The relationship between sleep duration, screen viewing time, and overweight/obesity was analyzed using a logistic regression analysis. Results The prevalence rates of overweight among boys and girls were 16.56% and 11.94%, respectively (χ2=13.59, P<0.05). The prevalence rates of obesity among boys and girls were 14.47% and 10.07%, respectively (χ2=14.01, P<0.05). In total, 74.41% students reported a lack of sleep; the average sleeping time was (9.24±1.07) h. The average sleeping time among boys was (9.35 ± 1.12) h and among girls was (9.13 ± 1.03) h. The difference in sleep duration between boys and girls was significant (t=5.79, P<0.05). The differences in sleep duration and overweight/obesity were significant between both boys (χ2=18.62, P<0.05) and girls (χ2=21.14, P<0.05). Regarding screen viewing time, 17.29% of students spent more than 2 hours per day viewing a screen. The difference in screen viewing time between boys and girls was significant (Z=3.02, P=0.014). The proportion of children with screen viewing time of more than 2 h/d among overweight/obese and healthy weight male students was 29.50% (82/278) and 22.56% (316/1401), respectively, which was significantly different (χ2=6.18, P=0.01). However, there was no significant difference when examining the same groups among girls (12.98% (24/185;obese/overweight) vs . 9.97% (136/1364; healthy weight); χ2=1.59, P=0.21). After adjusting for parental obesity, eating sweets, and physical activity, logistic regression analysis showed that students who had a sleep duration less than 10 h/d had an odds ratio of 1.4 (95% CI: 1.15-1.71), the odds ratio for boys and girls was 1.56 (95% CI: 1.13-2.14) and 1.69 (95% CI: 1.15-2.46). The students who had a screen viewing time of more than 2 h/d had an odds ratio of 1.14 (95% CI: 1.05-1.80); the odds ratio for boys in this group was 1.18 (95% CI: 1.03-1.67). Conclusion Short sleep duration is a risk factor for being overweight/obese in both boys and girls. However, long screen viewing times were associated with being overweight/obese only in boys.

Objective To observe the effect of arsenic exposure to drinking water on thelevel of histone 3 lysine 4 trimethylation (H3K4me3) and histone 3 lysine 79 trimethylation (H3K79me3) in peripheral blood leukocytes of human,and to analyze the relationship between arsenic exposure and H3K4me3,H3K79me3 modification levels.Methods A cluster sampling survey was carried out in typical endemic arsenicosis areas of Shanxi and Jilin provinces.Two hundred eighty-one local residents with a drinking water age of ≥ 10 years were selected as the survey subjects.According to the arsenic content of drinking water,the tested population was divided into control group (water arsenic content ≤0.01 mg/L,60 cases),low water arsenic exposure group (＞ 0.01-0.05 mg/L,61 cases),medium water arsenic exposure group (＞ 0.05-0.10 mg/L,50 cases),and 110 cases of high water arsenic exposure group (＞ 0.10 mg/L).Drinking water samples,immediate urine samples and peripheral blood samples were collected from the subjects.Arsenic content in drinking water and urinary arsenic content were determined via the atomic fluorescence method;histone H3K4me3 and H3K79me3 in peripheral blood leukocytes were determined by dot blot hybridization (Dot Blotting).Results There were no statistically significant differences in age (61.50,60.00,59.50,59.50 years old),different gender (male:20,27,17,40 cases,female:40,34,33,70 cases),body mass index (BMI),smoking and drinking status between the control group,low,medium and high water arsenic exposure groups.Water arsenic content in the control group,low,medium and high water arsenic exposure groups (median:0.005,0.024,0.076,0.150 mg/L),urinary arsenic content (0.011,0.018,0.061,0.134 mg/L),and water arsenic cumulative exposure levels (0.342,1.641,5.273,7.716 mg) were compared between groups,the differences were statistically significant (H =256.041,88.615,218.610,P ＜ 0.01).In the control group,low,medium and high water arsenic exposure groups,the modification levels of H3K4me3 (0.100,0.059,0.083,0.083)and H3K79me3 (0.049,0.036,0.055,0.052) in peripheral blood leukocytes were not significantly different (H =1.488,2.097,P ＞ 0.05).The levels of H3K4me3 and H3K79me3 in peripheral blood leukocytes were positively correlated with water arsenic content,urinary arsenic content,water arsenic cumulative exposure levels (r =0.245,0.221;0.299,0.318;0.149,0.149;P ＜ 0.01 or ＜ 0.05);there was a positive correlation between H3K4me3 and H3K79me3 modification levels (r =0.811,P ＜ 0.01).Conclusion There is a positive correlation between arsenic exposure through drinking water and the levels of H3K4me3 and H3K79me3 in the peripheral blood leukocytes of the population,but it is necessary to expand the sample size for further study.

Objective To investigate the diagnostic value of quantitative detection of α-synuclein and DJ-1 protein in saliva for Parkinson Disease. Methods Twenty seven patients diagnosed with primary Parkinson's disease and 27 healthy controls were studied.The clinical data of all subjects were collected.Each participant received a disease evaluation including Hohn-Yahr stage, unified Parkinson Disease Rating Scale (UPDRS)-Ⅱ / Ⅲ, 12 Item odor identification test from Sniffin'Sticks (SS-12), Montreal cognitive assessment (MoCA), Mini Mental State Examination (MMSE).α-syn and DJ-1 protein in saliva were examined by using enzyme linked immunosorbent assay (ELISA). The statistical analysis was used to test the difference in these two protein levels between patient and control groups and the correlation with age, gender and course of disease. Results There were significant changes in mean concentration of salivary α-syn (1269.02±16.09 pg/mL、1350.51±25.79 pg/mL,P=0.010) and DJ-1 protein (6.07±3.23 ng/mL、8.43±4.33 ng/mL,P=0.027) between patient and control groups. The sensitivity of α-syn and DJ-1 protein levels in PD diagnosis was 55.56% and 77.8%,and the specificity was 89.19% and 55.6%.The area under the ROC curve in finding the PD of α-syn and DJ-1 was 0.671 and 0.649, respectively. In Parkinson disease group, age, gender, UPDRS-Ⅱ / Ⅲ, Hohn-Yahr stage, SS-12, MMSE and MoCA of Parkinson's disease group were not related to the concentration of α-syn and DJ-1 protein in saliva (P>0.05). Conclusion The detection of α-syn and DJ-1 protein levels in saliva may be an auxiliary tool for diagnosis of PD.