Association of epicardial and visceral adipose tissue in relation to subclinical cardiac dysfunction in Chinese: Danyang study Authors: Fang Yao, Lingkai Zeng, Mulian Hua, Siqi Zhang, Junya Liang, Yun Gao, Chao Chen, Xixuan Zhao, Anxia He, Ming Liu Published: 2023 ABSTRACT ObjectiveOur study aims to examine the associations of visceral adipose tissue (VAT) and epicardial adipose tissue (EAT) with subclinical cardiac dysfunction in a Chinese population.DesignCross-sectional.BackgroundEAT and VAT are the most important ectopic fat pools which were previously shown to be associated with subclinical cardiac dysfunction. However, few studies simultaneously measured both EAT thickness and VAT area, and explored their associations with cardiac dysfunction. Our study aims to examine the associations of VAT and EAT with subclinical cardiac dysfunction in a Chinese population.MethodsThe study subjects were recruited from Danyang County from 2018 to 2019. Using Philips CX50, we recorded EAT thickness at the end-systole in a long-axis view. The subclinical systolic and diastolic function were assessed by two-dimensional speckle tracking, and transmitral and tissue Doppler imaging, respectively. Using Omron HDS-2000, we measured VAT area by dual bioelectrical impedance analysis.ResultsThe 1558 participants (age, 52.3±12.8 years) included 930 (59.7%) women. Compared with women, men had higher VAT area (99.4 vs 70.1 cm2; p<0.0001) but lower EAT thickness (4.02 vs 4.46 mm; p<0.0001). In simple correlation analyses, EAT thickness and VAT area were positively associated with E/e’ ratio (r=0.16 to 0.20; all p<0.0001) and negatively with global longitudinal strain (GLS) and e’ (r=−0.12 to −0.37; all p<0.0001). Furthermore, VAT area was associated with left ventricular ejection fraction (LVEF) (r=−0.14; p<0.0001). After adjustment for confounding factors, the association of EAT with GLS and that of VAT with e’ and E/e’ ratio remained significant (all p≤0.001), whereas the associations of EAT with subclinical diastolic dysfunction and that of VAT with systolic function became non-significant (all p≥0.11). Analyses on further adjustment for LVEF showed similar results.ConclusionsIncreased EAT thickness was associated with worse subclinical systolic dysfunction, while greater VAT area was associated with early diastolic dysfunction. Strengths and limitations of this study Both epicardial adipose tissue (EAT) thickness and visceral adipose tissue (VAT) area were measured, but did not measure waist circumference, which cannot be entirely substituted by VAT. Both epicardial adipose tissue (EAT) thickness and visceral adipose tissue (VAT) area were measured, but did not measure waist circumference, which cannot be entirely substituted by VAT. The relatively large sample (over 1000) of Chinese subjects with comprehensive information on medical history and blood biochemical measurement, but due to missing data and poor image quality, approximately 25% of participants were excluded in this study, which may have led to potential selection bias. The relatively large sample (over 1000) of Chinese subjects with comprehensive information on medical history and blood biochemical measurement, but due to missing data and poor image quality, approximately 25% of participants were excluded in this study, which may have led to potential selection bias. Subjects recruited in current study was population-based other than patients with various chronic disease Subjects recruited in current study was population-based other than patients with various chronic disease The design of our study was cross-sectional based, which does not enable us to relate cause and outcome. The design of our study was cross-sectional based, which does not enable us to relate cause and outcome. EAT thickness was evaluated using echocardiography but not MRI, the latter is considered as the golden standard for EAT measurement. EAT thickness was evaluated using echocardiography but not MRI, the latter is considered as the golden standard for EAT measurement. Introduction Obesity is a major global public health issue because of its high prevalence1 and detrimental effects on the development of cardiovascular disease. Our research2 and other studies3 4 have shown that obesity, as measured by body mass index (BMI), was closely related to subclinical abnormalities in cardiac structure and function that predispose to heart failure. Importantly, recent epidemiological studies have shown that ectopic fat localisation and its peculiar metabolic properties may be a better predictor for adverse cardiac function abnormalities than general obesity in obese individuals.5 6 These observations indicate that people have significantly different levels of cardiovascular risk even with similar body weight or BMI values.7 Ectopic adipose abundantly deposits in various parts of the body, such as the abdomen and epicardium cavity. The visceral adipose tissue (VAT) and epicardial adipose tissue (EAT) behave as not only anatomy barriers of the proximity organs but also metabolically active pools to secret adipokines and proinflammatory factors and are therefore considered to orchestrate an important regulatory network involving in systemic inflammation8 and eventually contributing to left ventricular (LV) functional alteration.9–11 However, most previous studies relating subclinical cardiac dysfunction to ectopic fat solely relied on either VAT12 13 or EAT measurement alone.14 15 Thus far, only three studies evaluated this relationship with both VAT and EAT in patients undergoing peritoneal dialysis,16 or after myocardial infarction17 or with suspected ischaemic heart disease18 but reported conflicting results. Whether the association between ectopic fat (VAT and EAT) and subclinical LV dysfunction, especially in population, remains to be determined. Therefore, in the current study, we investigated the correlation of VAT area and EAT thickness with subclinical systolic and diastolic dysfunction in a larger sample of the Chinese population. Materials and methods Study population As described previously,2 19 we recruited participants who were referred for comprehensive cardiovascular risk factors examination in Danyang Station Community, Jiangsu Province, China. We adhered to the principles of the Helsinki declaration and the Danyang epidemiological study protocol was approved by the Ethics Committee of Jiangsu Provincial Hospital of Chinese Medicine. All patients gave informed written consent. Of those enrolled in the period between June 2018 and May 2019, 1997 were eligible for inclusion criteria. A total of 439 subjects were excluded from this analysis because of missing information on visceral fat (n=6) and EAT measurements (n=252) or extreme values of echocardiographic examination (n=12) or poor quality of images (n=169). It was data of 1558 patients that completed the final analysis. Field work BMI was calculated as weight in kilograms divided by height in metres squared after body weight and height were determined as an experienced nurse measured. The same nurse measured each subject’s blood pressure (BP) three times consecutively by a validated Omron 7130 Oscillographic BP Monitor (Omron, Kyoto, Japan) during each interval the subject had rested for at least 5 min in the sitting position. The average of three BP readings was calculated for analysis. Hypertension was defined as a BP of at least 140 mm Hg systolic or 90 mm Hg diastolic, or as taking any anti-hypertensive drugs. The same observer also administered a standardised questionnaire to collect information on socio-demographic medical history and use of medications, lifestyle and physical activity. The Cambridge Index of physical activity was derived into four groups by combining occupational with recreational activity level: active, moderately active, moderately inactive and inactive.20 Overweight was defined as a BMI of 24 to less than 28 kg/m2, and obesity was defined as a BMI of at least 28 kg/m2 according to Chinese criteria.21 Coronary heart disease (CHD) was defined by undergoing percutaneous coronary intervention or coronary artery bypass grafting surgery, or by self-reported history of myocardial infarction. A trained nurse measured anthropometrics. Venous blood samples were drawn after overnight fasting for blood biochemical examination including glucose, lipid and renal function. Diabetes mellitus was defined as a fasting plasma glucose of at least 7.0 mmol/L or as taking anti-diabetic medicine. Dyslipidaemia was defined as a fasting Total Cholesterol (TC)≥6.2 mmol/L or low-density lipoprotein (LDL) cholesterol ≥4.1 mmol/L or triglycerides ≥2.3 mmol/L or HDL cholesterol <1.0 mmol/L or as the use of lipid lowering therapy.22 Echocardiographic measurements Comprehensive transthoracic echocardiographic measurements were performed with the Philips CX50 device (Philips, Bothell, Washington, USA), as recommended by the American Society of Echocardiography (ASE).23 We performed M-mode image to record LV chamber dimensions and wall thickness. Relative wall thickness (RWT) was the ratio of twofold posterior wall thickness to the LV end-diastolic dimension (LVEDD), and used as a measure of LV structure. LV mass was calculated using the formula recommended by ASE, and standardised to body size by dividing the LV mass by body surface area, as left ventricular mass index (LVMI). Left ventricular ejection fraction (LVEF) was calculated using the modified Simpson method. The early diastolic peak flow (E) and atrial peak flow (A) were measured by the pulse wave Doppler. Mitral annular e’ was measured in the septal and lateral mitral annulus by the tissue Doppler and was calculated as the mean of septal and lateral velocities. As previously described,24 speckle tracking analysis was evaluated offline using special software (QLAB Software V.9 Cardiac Motion/Mechanics Quantification, Philips). In brief, LV global longitudinal strain (GLS) was the mean value of the negative peak longitudinal strain from 17 ventricular segments obtained from the apical 4-chamber, 3-chamber and 2-chamber views. Lower absolute strain and e’ velocity values, as well as higher E/e’ ratio were used to indicate subclinical cardiac dysfunction. EAT was identified as the echo-free space between the outer wall of the myocardium and the visceral layer of pericardium in the parasternal long-axis section (figure 1).25 This was measured at end-systole in long-axis view perpendicular to the right ventricle free wall during three cardiac cycles and then averaged for analysis. Echocardiographic (A) and schematic demonstrating (B) epicardial adipose tissue, respectively. (A) Epicardial fat thickness (within red solid line) is identified as the echo-free space between the outer wall of the myocardium and the visceral layer of pericardium in the parasternal long-axis view. (B) Epicardial fat (within red dashed shape) between the myocardium and visceral layer of serous pericardium. Visceral adipose tissue measurement VAT area was measured using dual bioelectrical impedance analysis (BIA) with Omron HDS-2000 device (Omron, Kyoto, Japan) according to the protocol provided by the manufacturer. In brief, dual BIA calculates cross-sectional area of visceral fat at umbilical level using two electric paths to measure truncal impedance and surface impedance. The principles and formula have been described in detail elsewhere.26 27 Statistical analysis SAS software (V.9.4, SAS institute, Cary, North Carolina, USA) was used for database management and statistical analyses. Data are presented as means±SD or medians (25th and 75th percentiles) for continuous variables or as percentages for discrete variables, respectively. Comparisons of means, medians and proportions between groups relied on the Student’s t-test, Mann-Whitney U test and χ2 test, respectively. Single and multiple Pearson correlation analyses were applied to study the associations of EAT and VAT with subclinical cardiac systolic (LVEF and LV GLS) and diastolic dysfunction (e’ and E/e’ ratio). A two-sided value of p<0.05 was considered as statistically significant. Results Clinical and echocardiographic characteristics of the study participants The 1558 participants (52.3±12.8 years) included 930 (59.7%) women, 612 (39.3%) overweight, 149 (9.6%) obese patients, 604 (38.8%) patients with hypertension, 138 (8.9%) patients with diabetes and 493 (31.6%) patients with dyslipidaemia. Table 1 summarises the characteristics of the study participants. Most of the characteristics between men and women were similar, except that men had higher proportions of current smoking (49.8% vs 0.5%) and alcohol intake (36.0% vs 2.8%), more prone to be inactive, lower heart rate (−3.3 beats/min) and estimated glomerular filtration rate (−2.2 mL/(min×1.73 m2)) than women (all p≤0.005). In the aspect of metabolically parameters, compared with women, men had worse condition with greater BMI, higher BP, serum glucose, triglycerides, uric acid and lower HDL cholesterol (all p≤0.0002). Moreover, men had higher VAT area (99.4 vs 70.1 cm2; p<0.0001) but lower EAT thickness (4.02 vs 4.46 mm; p<0.0001). Table 1Demographic and laboratory characteristics by sexCharacteristicsTotal (n=1558)Men (n=628)Women (n=930)P valueAge, years52.3±12.851.6±13.852.7±12.20.11Body mass index, kg/m224.0±3.124.5±3.123.7±3.1<0.0001Systolic blood pressure, mm Hg127.0±18.7129.1±16.5125.6±19.90.0002Diastolic blood pressure, mm Hg80.8±10.482.9±10.079.3±10.4<0.0001Heart rate, beats per minute73.6±10.571.6±10.474.9±10.4<0.0001Current smokers, n (%)318 (20.4)313 (49.8)5 (0.5)<0.0001Alcohol drinkers, n (%)252 (16.2)226 (36.0)26 (2.8)<0.0001Obesity status<0.0001 Normal weight, n (%)797 (51.1)272 (43.3)525 (56.4) Overweight, n (%)612 (39.3)284 (45.2)328 (35.3) Obesity, n (%)149 (9.6)72 (11.5)77 (8.3) Hypertension, n (%)604 (38.8)258 (41.1)346 (37.2)0.12 Diabetes mellitus, n (%)138 (8.9)70 (11.6)68 (7.3)0.009 Dyslipidaemia, n (%)493 (31.6)243 (38.7)250 (26.9)<0.0001 Coronary heart disease, n (%)21 (1.4)10 (1.6)11 (1.2)0.49 Anti-hypertensive treatment, n (%)328 (21.1)116 (18.5)212 (22.8)0.04 Anti-hyperglycaemic treatment, n (%)99 (6.4)48 (7.64)51 (5.5)0.09 Lipid lowering therapy, n (%)24 (1.5)13 (2.1)11 (1.2)0.16Physical activity, n (%)<0.0001 Inactive357 (22.9)161 (25.6)196 (21.1) Moderately inactive870 (55.8)310 (49.4)560 (60.2) Moderately active296 (19.0)135 (21.5)161 (17.3) Active35 (2.3)22 (3.5)13 (1.4) Serum glucose, mmol/L5.26±1.185.42±1.295.15±1.08<0.0001 Total cholesterol, mmol/L4.87±0.934.80±0.944.92±0.930.01 HDL cholesterol, mmol/L1.44±0.321.35±0.291.51±0.32<0.0001 Triglycerides, mmol/L1.41 (0.98–2.12)1.51 (1.07–2.40)1.36 (0.95–1.98)<0.0001 Serum creatinine, µmol/L66.5±16.777.58±12.659.0±14.8<0.0001 Uric acid, µmol/L324.9±88.6375.8±83.5290.5±74.2<0.0001 eGFR, mL/(min×1.73 m2)98.8±15.097.5±15.299.7±14.80.005 VAT area, cm281.9±37.199.4±39.670.1±30.1<0.0001 EAT thickness, mm4.28±1.904.02±1.744.46±1.98<0.0001Data are mean±SD, median with IQR in parenthesis, or number with percentage in parenthesis. Definitions of overweight, obesity, hypertension, diabetes, dyslipidaemia and coronary heart disease are provided in Methods.HDL indicates high-density lipoprotein.EAT, epicardial adipose tissue; eGFR, estimated glomerular filtration rate; HDL, high-density lipoprotein; VAT, visceral adipose tissue. Demographic and laboratory characteristics by sex Data are mean±SD, median with IQR in parenthesis, or number with percentage in parenthesis. Definitions of overweight, obesity, hypertension, diabetes, dyslipidaemia and coronary heart disease are provided in Methods. HDL indicates high-density lipoprotein. EAT, epicardial adipose tissue; eGFR, estimated glomerular filtration rate; HDL, high-density lipoprotein; VAT, visceral adipose tissue. As shown in table 2, men and women had similar RWT and e’ (all p≥0.57). Of note, men had larger LVEDD (+2.7 mm), higher LVMI (+4.8 g/m2), lower LVEF (−0.8%) and LV GLS (−1.1%), but higher E/A ratio (+0.05), lower E peak (−7.0 cm/s) and E/e’ ratio (−0.8) (all p≤0.02). Table 2Echocardiographic characteristics by sexEchocardiographic measurementsTotal (n=1558)Male (n=628)Female (n=930)P valueLV structure LVEDD, mm45.8±3.847.4±3.644.7±3.5<0.0001 Relative wall thickness0.38±0.040.38±0.040.38±0.040.57 LV mass index, g/m279.5±14.382.4±14.177.6±14.1<0.0001LV systolic function LV ejection fraction, %66.9±3.266.4±3.367.2±3.1<0.0001 GLS, %20.7±2.120.1±1.921.2±2.1<0.0001LV diastolic function E peak, cm/s75.3±16.871.1±16.478.1±16.6<0.0001 E/A ratio1.07±0.371.10±0.401.05±0.350.02 e’, cm/s9.6±2.69.6±2.59.6±2.60.86 E/e’ ratio8.2±2.17.7±1.98.5±2.2<0.0001Data are mean±SD.GLS, global longitudinal strain; LV, left ventricular; LVEDD, left ventricular end diastolic dimension. Echocardiographic characteristics by sex Data are mean±SD. GLS, global longitudinal strain; LV, left ventricular; LVEDD, left ventricular end diastolic dimension. Association of epicardial and visceral adipose tissue with subclinical cardiac dysfunction We next investigated the relationship of EAT and VAT with various subclinical cardiac dysfunction. In simple Pearson correlation analyses, EAT (figure 2A–D) and VAT (figure 2E–2H) were positively associated with E/e’ ratio (r=0.16 to 0.20; all p<0.0001) and negatively with LV GLS and e’ (r=−0.12 to −0.37; all p<0.0001). Additionally, VAT area was significantly associated with LVEF (r=−0.14; p<0.0001). After adjustment for confounding factors including sex, age, BMI, systolic BP, current smoker and drinkers, physical activity, CHD, the use of anti-hypertensive, anti-hyperglycaemic and lipid-lowering agents, serum glucose, ratios of total to high-density lipoprotein cholesterol, uric acid, estimated glomerular filtration rate and LVMI, the associations of both EAT with subclinical systolic dysfunction (LV GLS) and VAT with subclinical diastolic dysfunction (e’ and E/e’ ratio) remained significant (all p≤0.001; table 3). However, the associations of EAT with subclinical diastolic dysfunction and that of VAT with systolic dysfunction (LVEF and GLS) became non-significant (all p≥0.11; table 3). Of note, after full adjustment for LVEF, the associations of EAT with GLS and that of VAT with e’ and E/e’ ratio were still statistically significant (all p≤0.0005; table 3). Table 3Correlations of epicardial and visceral adipose tissue in relation to subclinical cardiac dysfunction in all subjects (n=1558)Echocardiographic measurementsEAT thickness, mmVAT area, cm2Partial rP valuePartial rP valueLV systolic functionLV ejection fraction, % Adjusted model0.010.62−0.030.17 Fully adjusted model––––GLS, % Adjusted model−0.080.001−0.040.12 Fully adjusted model−0.090.0005−0.030.21LV diastolic functione’, cm/s Adjusted model−0.040.11−0.14<0.0001 Fully adjusted model−0.040.12−0.14<0.0001E/e’ ratio Adjusted model−0.010.750.100.0002 Fully adjusted model−0.010.730.100.0001Values are partial correlation coefficients (r). Covariates in adjusted model included sex, age, body mass index, systolic blood pressure, current smoker and drinkers, physical activity, coronary heart disease, the use of anti-hypertensive, anti-hyperglycaemic and lipid-lowering agents, serum glucose, ratios of total to high-density lipoprotein cholesterol, uric acid, estimated glomerular filtration rate and left ventricular mass index. Fully adjusted model additionally included LV ejection fraction, as appropriate.EAT, epicardial adipose tissue; GLS, global longitudinal strain; LV, left ventricular; VAT, visceral adipose tissue. Correlations of epicardial and visceral adipose tissue in relation to subclinical cardiac dysfunction in all subjects (n=1558) Values are partial correlation coefficients (r). Covariates in adjusted model included sex, age, body mass index, systolic blood pressure, current smoker and drinkers, physical activity, coronary heart disease, the use of anti-hypertensive, anti-hyperglycaemic and lipid-lowering agents, serum glucose, ratios of total to high-density lipoprotein cholesterol, uric acid, estimated glomerular filtration rate and left ventricular mass index. Fully adjusted model additionally included LV ejection fraction, as appropriate. EAT, epicardial adipose tissue; GLS, global longitudinal strain; LV, left ventricular; VAT, visceral adipose tissue. Simple correlation of epicardial fat thickness (left panel) and visceral fat area (right panel) with systolic and diastolic function measures. (A) Left ventricular ejection fraction; (B) left ventricular global longitudinal strain; (C) e’; and (D) E/e’ ratio; (E) left ventricular ejection fraction; (F) left ventricular global longitudinal strain; (G) e’; and (H) E/e’ ratio. The correlation coefficient r and p value for simple Pearson correlation analyses between visceral fat area and cardiac function are given, respectively. Regression line (solid line) and 95% CIs (dotted lines) are also given, respectively. Discussion Our study for the first time revealed that EAT thickness is negatively associated with subclinical LV systolic dysfunction as assessed by GLS but not diastolic dysfunction. In contrast, greater VAT area is associated with worse subclinical diastolic dysfunction, as reflected by e’ and E/e’ ratio. As decreased GLS, lower e’ and higher E/e’ may be predictive of subsequent clinical heart failure,28 our findings therefore may have important clinical significance for cardiovascular disease prevention by modulating the epicardial and visceral fat depot in addition to controlling body weight. Recently, a published position statement on visceral obesity29 recommended that ectopic fat, such as VAT or EAT, should be assessed in addition to body weight because emerging evidences suggest ectopic fat as novel risk factors for heart failure.9 30 However, most studies12–15 explored associations of subclinical cardiac dysfunction with either VAT or EAT. We simultaneously measured VAT area, EAT thickness e’, E/e’ ratio and GLS in a much larger population-based sample, which provided the possibility to find weak and context-dependent associations. Our findings therefore offer clarity to the relationship between ectopic fat and subclinical cardiac dysfunction, and provide strong support for its potential clinical application in evaluating risk for cardiovascular disease. Thus far, only three studies with a small sample size investigated the associations of EAT and VAT in relation to subclinical diastolic dysfunction. In 149 patients with peritoneal dialysis, Lin et al16 measured EAT with echocardiography and reported a similar thickness of EAT (3.80 mm) as our study. However, in this previous study, EAT rather than VAT was dramatically related with early diastolic dysfunction assessed by E/e’. In 255 patients with myocardial infarction, Fontes-Carvalho et al17 assessed epicardial fat volume and VAT area by CT scan and discovered epicardial fat volume was associated with early diastolic dysfunction. The discrepancies in association of EAT with cardiac diastolic dysfunction between our study and that of these two studies16 17 are supposed due owing to numerous core factors including subjects’ characteristics, sample size and methods for EAT measurement. First, in comparison to a small number of patients with predefined disease, our study with 1558 subjects was population-based, thus having high statistical power to provide the possibility of finding context-dependent associations. Second, methods for EAT assessment were different (echocardiography vs CT) between our study and Fontes-Carvalho et al. Similar to our study, Takahari et al18 found that independent of BMI, VAT area was associated with e’. However, in this previous study, VAT area was not associated with E/e’. This study apparently has insufficient power to show association of the size observed in our study. However, up to now, no previous study had concurrently measured VAT and EAT and explored their associations with subclinical systolic dysfunction. Our finding on the association between EAT and GLS is consistent with the results of previous studies.15 31 The underlying mechanism might be that EAT may increase systemic inflammatory response and further impair GLS by promoting myocardial fibrosis.8 Nonetheless, we did not find a significant association between VAT and GLS, which was in conflict with a previous study.32 This discrepancy is probably attributed to BMI adjustment. Indeed, significant association relating VAT with GLS was also observed when we did not adjust for BMI (data not shown). Strengths and limitations We performed both EAT thickness and VAT area in 1558 Chinese participants, which made it possible to explore the potential differential impacts of these two adipose tissues on subclinical cardiac systolic and diastolic dysfunction. Additionally, in contrast to previous studies,16–18 participants recruited in our study were population-based other than patients with various chronic diseases. Despite these strengths, our study should be interpreted within the context of its potential limitations. A main limitation is that the design nature of our study is cross-sectional based, which does not enable us to relate cause and outcome. Future prospective study is warranted to explore the cause inference. Another limitation of this study is that we only measured EAT thickness with echocardiography but not MRI in the current study. The latter is considered as the golden standard for EAT measurement. Nonetheless, one recent study has reported that EAT thickness (assessed by echocardiography) was well related with total EAT mass (measured by MRI).33 Due to missing data and poor image quality, approximately 25% of participants were excluded in this study, which may have led to potential selection bias. In future studies, we will strengthen data management and focus on data completeness to enhance data reliability and minimise exclusions that could introduce bias. Finally, central obesity stands as one of the predictive factors for subclinical LV diastolic dysfunction.34 However, our study did not measure waist circumference, which cannot be entirely substituted by VAT, constituting a limitation in this research. In future studies, we aim to supplement the inclusion of waist circumference as crucial data for measurement. This will be integrated into the statistical model for analysis and comparison with VAT. Conclusions In summary, we revealed that ectopic adipose depot location is associated with subclinical differences in systolic or diastolic function in a Chinese population. Higher EAT thickness was associated with worse subclinical systolic dysfunction, whereas greater VAT area was associated with early diastolic dysfunction. Further prospective and interventional studies are needed to determine whether reducing the mass of EAT or VAT can improve cardiac function and prevent the development of heart failure. Supplementary Material Reviewer comments Author's manuscript