Linear Regression Adjustment Research Articles

Abstract P4-07-21: Neo-adjuvant chemotherapy in various breast cancer subtypes in the 21st century

Abstract Introduction Breast cancer is a complex, heterogeneous disease encompassing a spectrum of subtypes with distinct biological features, each having unique responses to various treatment modalities and different clinical outcomes. The type of neoadjuvant chemotherapy (NAC) for locally advanced breast cancer is decided based on breast cancer subtype, which includes triple negative breast cancer (TNBC), hormone receptor positive breast cancer (HR+BC), HER2+ breast cancer (HER2+BC) and triple positive breast cancer (TPBC). Multiple NAC regimens exist but have not been directly compared to determine the optimal treatment regimen in patients with various stages (I-III) and subtypes of breast cancer. The objective of this study was to assess pathological complete response (pCR) rates in patients treated with various types of NAC and analyze associated clinical factors in our diverse patient population. Methods This study included 297 patients treated with NAC for breast cancer between 2015 and 2021 at LCMC Heath, New Orleans, Louisiana. The tumor, lymph node, metastasis (TNM) system was used for clinical and pathological staging. Biologic subclassification using estrogen receptor (ER), progesterone receptor (PR), HER2 were performed. Response to NAC was documented as pCR when there was no evidence of residual invasive tumor in the breast or axillary lymph nodes. Categorical variables were summarized by reporting counts and percentages, while continuous variables were summarized by reporting means and standard deviations. Fisher exact tests were used to compare pCR status by each chemotherapy status or demographic factor. Wilcoxon Rank-sum tests were used to compare continuous variables across pCR groups. Multivariable linear regression was performed to predict overall tumor shrinkage %. Results Among all patients, median age was 54.75 years (min-max: 22-78). 30 (10%) patients were stage I, 175 (58.9%) were stage II, 89 (29.9%) were stage III and 3 (1%) were stage IV with oligometastatic disease. 171 (57.6%) were African American (AA), 82 (27.6%) were Caucasians and 41(13.8%) were other race. Table 1. shows chemotherapy and pCR rates among different breast cancer subtypes. Patients with carboplatin (57.3 vs 35.3; p< 0.001), had a higher likelihood of complete remission than non-users. After multivariable linear regression adjustment, Paclitaxel increased the % reduction in tumor size significantly (EST= 39%, CI 2% – 77%; p=0.042) compared to non-users. In the subset of TNBC patients this held in terms of pCR rates (47.1% vs 25%, p=0.036). In TPBC, pCR was higher in younger (p=0.028) and non-AA (p=0.0023) patients. Conclusion Multiple NAC regimens for breast cancer exist and optimization of regimens is key. We explored the use of several different chemotherapy agents and found the use of carboplatin beneficial, while doxorubicin, cyclophosphamide and cisplatin had a decreased likelihood of achieving a pCR; however, this may be due to the intrinsic nature of the subtypes that would be treated with these NAC regimens. We plan to explore by subtype and treatment regimen in future analysis.Interestingly, AA have significantly less pCR in TPBC, compared to non-AA, although this finding was not seen in other subtypes. Future studies are needed to investigate this further. Table 1: Chemotheraphy and pCR rates among different breast cancer subtypes. Citation Format: Aneesha Ananthula, Katharine Thomas, Lily Chen, Vincent K. Carey, Caitlin M. Sullivan, Karla M. Martin, Emily Baas, Kara C. Plasko, Brandi Sun, Madison Lanza, Cindy Nguyen, Andrew Chapple, Michelle M. Loch. Neo-adjuvant chemotherapy in various breast cancer subtypes in the 21st century. [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr P4-07-21.

Analysis of the factors associated with agricultural GDP in Mexico

Objective: To detect factors of production which can explain the total value of agricultural production (agricultural GDP) and to carry out a descriptive analysis. Methodology: A multiple linear regression adjustment and a descriptive analysis of the variables were carried out, using the available information. Results: Five variables have a significant association with agricultural GDP. In the last 20 years, the agricultural sector has experienced an exponential economic growth, with a 7.7% annual growth rate (AAGR). About 60% of the total production value is provided only by 7 states. Agricultural exports have increased from 2015, as a result of the USA demands for Mexican fruits and vegetables. Implications: The information available about the state of agriculture in Mexico makes it possible to understand trends and take actions to increase the agricultural GDP. The analysis is limited by the lack of information about many variables. Conclusions: Sown areas have decreased and the factors of production analyzed maintain a positive average annual growth rate, which indicates a process of transition from extensive crops (such as grains and cereals) to intensively exploited crops (such as fruits and vegetables). An investment aimed to generate agricultural data is needed

DIRECTIONS OF LINE CONSTRAINTS DILEMMA IN PARAMETRIC LINEAR REGRESSION

The demand for positional accuracy and multi-dimensional data have demonstrated drastic changes in the geomatics data adjustment approach. Furthermore, the capability of modern sensors to provide high accuracy data (i.e., global navigation satellite system) has caused the crucial requirement for a rigorous adjustment that can process data from multi-sensors. Geomatics practitioners have gradually transformed the adjustment procedure to the most rigorous approach (i.e., parametric linear regression) to adapt to current demand. However, legacy datasets that utilize independent line constraint in the traditional adjustment approach have caused significant uncertainties in parametric linear regression (LR) adjustment. To resolve this dilemma, this research has designed robust experiments using closed traverse types: single-line constraint, multi-line constraints, and sub-network line constraint. Through errors trend and network form deterioration analyses, the outcomes have visually and numerically verified the insignificant of independent line constraints in parametric LR. However, the establishment of control points at the beginning or end of lines could solve the limitation of the abovementioned issue. In both analyses, control points at initial lines have demonstrated the best solution for constrained adjustment. The obtained results have exemplified the appropriate implementation of network adjustment in the presence of line constraints. As positional accuracy becomes the main priority, it can be concluded that points-based constraints are more advisable in preserving the quality of cadastral network adjustment.

Automatic algorithm for quantifying lung involvement in patients with chronic obstructive pulmonary disease, infection with SARS-CoV-2, paracoccidioidomycosis and no lung disease patients.

In this work, we aimed to develop an automatic algorithm for the quantification of total volume and lung impairments in four different diseases. The quantification was completely automatic based upon high resolution computed tomography exams. The algorithm was capable of measuring volume and differentiating pulmonary involvement including inflammatory process and fibrosis, emphysema, and ground-glass opacities. The algorithm classifies the percentage of each pulmonary involvement when compared to the entire lung volume. Our algorithm was applied to four different patients groups: no lung disease patients, patients diagnosed with SARS-CoV-2, patients with chronic obstructive pulmonary disease, and patients with paracoccidioidomycosis. The quantification results were compared with a semi-automatic algorithm previously validated. Results confirmed that the automatic approach has a good agreement with the semi-automatic. Bland-Altman (B&A) demonstrated a low dispersion when comparing total lung volume, and also when comparing each lung impairment individually. Linear regression adjustment achieved an R value of 0.81 when comparing total lung volume between both methods. Our approach provides a reliable quantification process for physicians, thus impairments measurements contributes to support prognostic decisions in important lung diseases including the infection of SARS-CoV-2.

Lipidomics-Based Comparison of Molecular Compositions of Green, Yellow, and Red Bell Peppers.

Identifying and annotating the molecular composition of individual foods will improve scientific understanding of how foods impact human health and how much variation exists in the molecular composition of foods of the same species. The complexity of this task includes distinct varieties and variations in natural occurring pigments of foods. Lipidomics, a sub-field of metabolomics, has emerged as an effective tool to help decipher the molecular composition of foods. For this proof-of-principle research, we determined the lipidomic profiles of green, yellow and red bell peppers (Capsicum annuum) using liquid chromatography mass spectrometry and a novel tool for automated annotation of compounds following database searches. Among 23 samples analyzed from 6 peppers (2 green, 1 yellow, and 3 red), over 8000 lipid compounds were detected with 315 compounds (106 annotated) found in all three colors. Assessments of relationships between these compounds and pepper color, using linear mixed effects regression and false discovery rate (<0.05) statistical adjustment, revealed 11 compounds differing by color. The compound most strongly associated with color was the carotenoid, β-cryptoxanthin (p-value = 7.4 × 10−5; FDR adjusted p-value = 0.0080). These results support lipidomics as a viable analytical technique to identify molecular compounds that can be used for unique characterization of foods.

Association of vitamin D nutritional status with body muscle mass in school-age children adolescents

Objective: To investigate the association between vitamin D nutritional status and the body muscle mass in children. Methods: Data were obtained from School-based Cardiovascular and Bone Health Promotion Program. In 2017, a total of 15 391 children aged 6-16 years in Beijing were selected through stratified cluster sampling in baseline survey. A follow-up investigation was conducted in 2019. The questionnaire survey and the detection of serum 25-hydroxyvitamin D [25(OH)D] level were conducted. The bioelectrical impedance analysis (BIA) apparatus was used to measure body muscle mass, and muscle mass index (MMI) was calculated. Multivariable linear models were used to analyze the association of vitamin D nutritional status with the baseline and follow-up MMI measures. Results: A total of 10 890 children aged (11.5±3.3) years(boys accounting for 49.6%) were included in the analysis. The average 25(OH)D level was (35.4±12.0) nmol/L, with an adequacy ratio of 11.1%. After multivariate linear regression adjustment for age, sex, body fat mass, smoking status, alcohol use status, dairy supplement, calcium supplement, physical activity, and pubertal development, no statistically significant association between vitamin D nutritional status and baseline MMI level was observed (P>0.05). For the follow-up MMI, the Z-score increased by 0.008 (P=0.058) for per 10 nmol/L increase in 25(OH)D, which were 0.002 (P=0.815) and 0.037 (P=0.031) higher in children with insufficient and adequate vitamin D than those with vitamin D deficiency, respectively (P for trend =0.089). Subgroup analysis showed that in the normal BMI group, for per 10 nmol/L increase in 25 (OH) D, the MMI at baseline survey and MMI Z-score at follow-up of children with adequate vitamin D and increased by 0.019 and 0.014, respectively (both P<0.05). Conclusions: Vitamin D nutritional status was related to muscle mass in children, and children with adequate vitamin D tended to obtain higher MMI. Children and adolescents are encouraged to maintain sufficient vitamin D levels, strengthen nutrition and exercise to promote body health.

Local Linear Forests

Random forests are a powerful method for nonparametric regression, but are limited in their ability to fit smooth signals. Taking the perspective of random forests as an adaptive kernel method, we pair the forest kernel with a local linear regression adjustment to better capture smoothness. The resulting procedure, local linear forests, enables us to improve on asymptotic rates of convergence for random forests with smooth signals, and provides substantial gains in accuracy on both real and simulated data. We prove a central limit theorem valid under regularity conditions on the forest and smoothness constraints, and propose a computationally efficient construction for confidence intervals. Moving to a causal inference application, we discuss the merits of local regression adjustments for heterogeneous treatment effect estimation, and give an example on a dataset exploring the effect word choice has on attitudes to the social safety net. Last, we include simulation results on real and generated data. A software implementation is available in the R package grf. Supplementary materials for this article are available online.

Stochastic One-Step Training for Feedforward Artificial Neural Networks

This paper studies the use and application of a fast method (non-iterative and instantaneous) for Feedforward Neural Networks training in which the weights of the hidden layer are assigned randomly, and the weights of the output layer are trained through a linear regression adjustment. The method solves two of the problems that are present in traditional training: training time and optimal structure. While traditional iterative training methods require long periods to train a single structure, the proposed method allows training a structure in a single step (not iterative). In this way, by scanning the number of neurons in the hidden layer, many structures are trained in a short time, and it is possible to obtain an optimal topology. A quality control criterion of the predictions is proposed based on the coefficient of determination that guarantees short times and an optimal number of hidden neurons to characterize a specific problem. The feasibility of the proposed method is tested by comparing its performance against building functions of the artificial neural networks toolbox in Matlab®, resulting superior in both approximation quality and training time. A rigorous study and analysis are performed for the regression of simulated data on two different surfaces with a specific noise and different topologies of the neural network. The resulting process time is at least 150 times shorter for proposed training than with the iterative training that Matlab uses, thus obtaining well-founded learning rules. A novel way of an amputated matrix is proposed that breaks the paradigm of the way multiple-output systems are trained and improves the quality of predictions with no detriment to training times.

Healthy Days at home: A novel population-based outcome measure

BackgroundHealthy Days at Home (HDAH) is a novel population-based outcome measure developed in conjunction with the Medicare Payment Advisory Commission. MethodsWe identified beneficiary age, sex, race, and Medicaid eligibility, death date, chronic conditions and healthcare utilization among a 20% sample of Medicare beneficiaries in 2016. For each beneficiary we calculated HDAH for the year by subtracting the following measure components from 365 days: mortality days, the total number of days spent in inpatient, observation, skilled nursing facilities (SNF), inpatient psychiatry, inpatient rehabilitation and long-term hospital settings as well as the number of outpatient emergency department and home health visits. We examined how HDAH and its components varied by beneficiary demographic characteristics and chronic condition burden as well as by healthcare market (Hospital Referral Region). We specified a patient-level linear regression adjustment model with HDAH as the outcome and incorporated market fixed effects as well as beneficiary age, sex, and Chronic Conditions Warehouse categories as covariates. We examined the impact of including home health visits in the measure, as well as the association between market demographics and health system characteristics and mean market HDAH. We examined how HDAH changed from 2013 to 2016. ResultsThe 6,637,568 beneficiaries age 65 and older in our sample had a mean of 347.2 HDAH, those 80 and older had a mean of 325.3 while those with three or more chronic conditions had a mean of 333.7. The components that led to the largest reduction in HDAH were mortality (7.4 days), home health (2.7 visits), SNF utilization (2.4 days) and inpatient care (1.5 days). The worst performing market had 5.8 fewer adjusted HDAH on average compared to the national mean, while beneficiaries in the best-performing market had 5.0 more HDAH on average compared to the national mean, among all beneficiaries age 65 and older. ConclusionsHDAH is a population-based quality measure with substantial market-level variation. ImplicationsHDAH recognizes the multidimensional nature of healthcare and may afford providers greater flexibility to tailor quality-improvement initiatives to the unique needs of their patients. Level of evidenceLevel II.

PC-FACS

PC-FACS

The Magnitude and Direction of Collider Bias for Binary Variables

Abstract Suppose we are interested in the effect of variable X on variable Y. If X and Y both influence, or are associated with variables that influence, a common outcome, called a collider, then conditioning on the collider (or on a variable influenced by the collider – its “child”) induces a spurious association between X and Y, which is known as collider bias. Characterizing the magnitude and direction of collider bias is crucial for understanding the implications of selection bias and for adjudicating decisions about whether to control for variables that are known to be associated with both exposure and outcome but could be either confounders or colliders. Considering a class of situations where all variables are binary, and where X and Y either are, or are respectively influenced by, two marginally independent causes of a collider, we derive collider bias that results from (i) conditioning on specific levels of the collider or its child (on the covariance, risk difference, and in two cases odds ratio, scales), or (ii) linear regression adjustment for, the collider or its child. We also derive simple conditions that determine the sign of such bias.

Temporal analysis of the relationship between leptospirosis, rainfall levels and seasonality, Santa Catarina, Brazil, 2005-2015.

ABSTRACTThe objective of the study is to analyze the temporal trend of leptospirosis incidence, according to rainfall levels in Santa Catarina, according to seasonality, from 2005 to 2015. This is an ecological study of time series, with date of leptospirosis, rainfall levels and population. The incidence rates of leptospirosis, relative excess of incidence, Pearson's correlation coefficient (r) and an angular coefficient (β) were analyzed from the linear regression adjustment, with a 5% significance level. Distribution of leptospirosis cases, rainfall levels and cases reason/rainfall levels, stratified by month of occurrence were presented. There were 5,274 cases of leptospirosis, with the monthly average being 439 cases, ranging from 211 in September to 770 in January. The mean rate of leptospirosis was 7.03 per 100,000 habitants. The average rainfall level was 158.68 mm, with the lowest levels occurring in August, average of 124.9 mm, and the highest in January average of 213.20 mm. The positive correlation between leptospirosis rates and rainfall levels, during the period from January to December (r = 0.68, p = 0.023), indicates a positive temporal association between the amount of rainfall and the cases of the disease. The disease occurred all year round and presented a distinct seasonality from October to March.

The elastic capacity of a tendon-repair construct influences the force necessary to induce gapping.

Most biomechanical investigations of tendon repairs were based on output measures from hydraulic loading machines, therefore, accounting for construct failure rather than true gapping within the rupture zone. It was hypothesized that the elastic capacity of a tendon-repair construct influences the force necessary to induce gapping. A tendon-repair model was created in 48 porcine lower hind limbs, which were allocated to three fixation techniques: (1) Krackow, (2) transosseous and (3) anchor fixation. Loading was performed based on a standardized phased load-to-failure protocol using a servohydraulic mechanical testing system MTS (Zwick Roell, Ulm, Germany). Rupture-zone dehiscence was measured with an external motion capture device. Factors influencing dehiscence formation was determined using a linear regression model and adjustment performed as necessary. A 3-mm gap was considered clinically relevant. Analysis of variance (ANOVA) was used for comparison between groups. The elastic capacity of a tendon-repair construct influences the force necessary to induce gapping of 3mm (F3mm) [β = 0.6, confidence interval (CI) 0.4-1.0, p < 0.001]. Furthermore, the three methods of fixation did not differ significantly in terms of maximum force to failure (n.s) or F3mm (n.s). The main finding of this study demonstrated that the higher the elastic capacity of a tendon-repair construct, the higher the force necessary to induce clinically relevant gapping. Controlled biomechanical study.

Spatial and temporal variability of spontaneous grass cover and its influence on sediment losses in an extensive olive orchard catchment

Vegetal covers protect the soil against erosion in agricultural systems. In marginal olive farms, spontaneous grass covers are preferred due to their low cost, despite their high spatial and temporal heterogeneity which limits their efficiency. Although the impact of different managements, soil conditions or cover crops species has been evaluated and compared, there is currently insufficient information available on the spatial and temporal patterns of spontaneous grass covers in olive groves. This work focuses on the analysis of the spatial distribution patterns, temporal stability and protection capacity of spontaneous grass covers in olive orchards. First, the temporal variability of spontaneous grass cover was explored in relation to the meteorological conditions and the management system. Second, spatial analysis by distance indices (SADIE) was applied to explore the spatial patterns and persistence of the spontaneous cover. Finally, a descriptive analysis of events was made to illustrate the degree of influence of the cover on the hydrological response of the catchment.A total 12 surveys were carried out on a seasonal scale over 3years (2011−2013) to measure the cover percentage of spontaneous grass through surface photograph analyses. A regular, permanent grid of 37 points (6samples/ha) was used to define each sample-location. In order to evaluate the relations between the cover percentage and meteorological variables, multiple linear regressions were adjusted while the SADIE approach and correlation analysis were used to describe possible spatial aggregation patterns and its dependence on topographical and soil attributes such as aspect, slope, drainage area, height, richness of grass species and apparent electrical conductivity.The average annual percentage of cover ranged from 23% to 36%, with a coefficient of variation of 57% and 6%, respectively. On the seasonal scale, the cover varied between 0.2% and 50%. Accumulated precipitation during the precedent 15days, the average of minimum temperature of the previous 60days and the cumulated potential evapotranspiration of the same period showed a significant correlation with the mean cover percentage in the catchment. A simple multiple linear regression adjustment between the cumulated precipitation for the 15 previous days and the number of months since the vegetation was removed using herbicide or by summer conditions gave a determination coefficient of 77%. Moreover, a permanent spatial pattern was observed for periods characterized by abundant preceding rainfall. Only the apparent electrical conductivity of the topsoil (50cm in depth) showed significant correlations with the spatial patterns of spontaneous grass. Finally, similar rainfall events taking place with different degrees of soil cover presented very different hydrological responses, which allowed us to quantify the contribution of spontaneous vegetation to sediment dynamics during humid periods.

A Nonparametric Bayesian Analysis of Heterogenous Treatment Effects in Digital Experimentation

Randomized controlled trials play an important role in how Internet companies predict the impact of policy decisions and product changes. In these “digital experiments,” different units (people, devices, products) respond differently to the treatment. This article presents a fast and scalable Bayesian nonparametric analysis of such heterogenous treatment effects and their measurement in relation to observable covariates. New results and algorithms are provided for quantifying the uncertainty associated with treatment effect measurement via both linear projections and nonlinear regression trees (CART and random forests). For linear projections, our inference strategy leads to results that are mostly in agreement with those from the frequentist literature. We find that linear regression adjustment of treatment effect averages (i.e., post-stratification) can provide some variance reduction, but that this reduction will be vanishingly small in the low-signal and large-sample setting of digital experiments. For regression trees, we provide uncertainty quantification for the machine learning algorithms that are commonly applied in tree-fitting. We argue that practitioners should look to ensembles of trees (forests) rather than individual trees in their analysis. The ideas are applied on and illustrated through an example experiment involving 21 million unique users of EBay.com.

Health-related quality of life and burden of fatigue in patients with primary immune thrombocytopenia by phase of disease.

The main objective of this study was to compare health-related quality of life (HRQOL) of primary immune thrombocytopenia (pITP) patients with that of general population, overall, and by patient group (i.e., newly diagnosed, persistent, and chronic patients). Fatigue was also investigated as a secondary objective. Overall, 424 adult patients were enrolled in a multicenter observational study and the control group consisted of a representative sample from the general population. Propensity score matching plus further multivariate linear regression adjustment was used to compare HRQOL outcomes between pITP patients and general population. Mean age of patients was 54years. Of those with HRQOL assessment, 99 patients (23.6%) were newly diagnosed, 53 (12.6%) were persistent, and 268 (63.8%) were chronic pITP patients. Comparison by patient group versus their respective peers in the general population revealed greater impairments in persistent pITP patients. Persistent pITP patients reported clinically meaningful impairments in physical functioning (-15; 95% CI -24.1 to -5.8; P=0.002), social functioning (-15.3; 95% CI -25.5 to -5.1; P=0.004), role physical (-28.4; 95% CI -43.1 to -13.7; P<0.001), role emotional (-23.9; 95% CI -40.1 to -7.7; P=0.004), and mental health scales (-11.3; 95% CI -21.2 to -1.4; P=0.026) of the SF-36 questionnaire. Higher fatigue severity was associated with lower physical and mental HRQOL outcomes. Our findings suggest that the burden of the disease and treatment might depend on the disease phase and that persistent pITP patients are the most vulnerable subgroup. Am. J. Hematol. 91:995-1001, 2016. © 2016 Wiley Periodicals, Inc.

Analysis of metabolic pools in broilers chicks

This paper shows the possibility of obtaining new parameters for the mathematical modelling of data on stable isotopes in biological systems and its application in obtaining data on metabolic pools of blood plasma, blood serum, liver and muscle of broilers. This theory states that the modelling of turnover used for studies of isotopic incorporation when the metabolism has a single metabolic pool is feasible by the technique of setting an exponential. However, when the metabolism has more than one metabolic pool, it is necessary to apply the linearization technique, linear regression adjustment and evaluation of the assumptions of regression to obtain the kinetic parameters such as half-life (T1/2) and isotope exchange rate (k). The application of this technique on carbon-13 data from 100 one-day-old chicks, with the change of diet composed of grains of the photosynthetic cycle of plants from C4 to C3, in broilers has enabled the discovery that the liver, blood plasma and blood serum have a single metabolic pool; however, the pectoral muscle has two metabolic pools. For the liver, blood plasma and blood serum, the half-life values were found by the exponential fit being T1/2 = 1.4 days with the rate of exchange of k = 0.502, T1/2 = 2.4 days with k = 0.293 and T1/2 = 2.0 days with k = 0.348, respectively. For the pectoral muscle, after linearization, the half-life values were found for T1/2(1) = 1.7 and T1/2(2) = 3 days, with exchange rates of k1 = 0.405 and k2 = 0.235, representing approximately 66 and 34 %, respectively.

Skin pigmentation interferes with the clinical measurement of regional cerebral oxygen saturation

Devices utilizing near-infrared (NIR) spectroscopy have been used to assess regional intracerebral oxygen saturation (rSO2) during anaesthesia for a decade. The presence of wide differences among individuals reduces their applicability to steady-state measurements. Current devices may not adequately account for variations in skin pigmentation. From our ongoing departmental registry, 3282 consecutive patients underwent cardiac surgery between 2010 and 2012 and their pre-induction measurements of rSO2 were available. Of these, 2096 identified themselves as Caucasian (Cauc) and 1186 as African-American (AA). Pre-induction rSO2, clinical and operative features were compared. Clinical and operative details of these patients differed widely between the two populations. High-risk features were more common in AA patients, but no difference in mortality was observed (4.8% in AAs vs 4.7% in Caucs, P=0.87). Preprocedure rSO2 was systematically higher in Cauc (65.5% vs 53.3%, P<0.001). After multivariate linear regression adjustment, AA ethnicity proved to be associated independently with low rSO2 [odds ratio (OR) -8.28, 95% confidence interval (CI) -9.12 to -7.44, P<0.001]. Multivariate logistic regression analysis showed that preprocedural rSO2 was independently associated with operative mortality both in the Cauc group (OR 0.97, 95% CI 0.96-0.99, P=0.001) and in the AA group (OR 0.97, 95% CI 0.95-0.99, P=0.01). AAs have a lower rSO2 than Caucs as measured by the INVOS 5100C cerebral oximeter. Reasonably, this could be attributed to attenuation of the NIR light by skin pigment. Despite this limitation, in both ethnic groups, lower preoperative rSO2 was predictive of greater operative mortality.

An improved SCGM(1,m) model for multi-point deformation analysis

Considering the deformation of discrete monitoring points within the same deformable body usually have similar physical properties and tend to undergoing identical dynamic process, joint modelling of the deformation processes of these points in time domain are expected to generate better results. Yin et al. (1997) first extended the multi-variable grey model-system cloud grey model SCGM(1,m), with obviously superior modelling mechanism than single-variable grey model, to multi-point deformation modelling. However, this model is still not widely recognized and its applications remain very limited in the field of deformation analysis. The objective of this study is to demonstrate the capability of the SCGM(1,m) model, to present two revisions to further improve the performance of the model and to draw more attention to the community of deformation analysis. We first introduce the principles of the SCGM(1,m) model in the analysis and prediction of deformation surveys. Two practical techniques, namely residuals re-modelling and linear regression adjustment, are then presented to improve the SCGM(1,m) model. Combined with slope monitoring data, the modelling with the original and the improved SCGM(1,m) models by residuals re-modelling and linear regression adjustment are illustrated. The mean relative prediction errors decrease from 5.89% to 3.54% and 2.69%, when the two refining techniques are applied, respectively, indicating relative improvements of 39.9% and 54.3%.

Elevated Levels of Macrophage Migration Inhibitory Factor in Women with Metabolic Syndrome

Metabolic syndrome is a complex clinical disorder characterized by obesity, a disturbance of glucose metabolism, dyslipidemia, and hypertension, leading to increased cardiovascular risk. Macrophage migration inhibitory factor (MIF) is a proinflammatory cytokine produced both by innate immune cells and by adipocytes, and it plays an important role in inflammatory and cardiovascular diseases. The goal of this study was to evaluate the expression of circulating MIF in patients with metabolic syndrome. A study was conducted involving 172 persons who attended the Jeju National University Hospital Health Promotion Center. Among the 172 subjects, 88 patients with metabolic syndrome and 84 healthy control subjects were included. Serum MIF levels were considerably higher in patients with metabolic syndrome than in healthy subjects (mean±SEM, 1413.0-pg/ml±102.6 vs. 1077.0-pg/ml±-91.3, p=0.016). Among the metabolic syndrome patients, MIF levels were significantly increased in women (1403.0-pg/ml±114.2 vs. 921.3 pg/ml±117.3, p=0.005), but not in men. Even after further linear regression adjustment for age and body mass index, the expression of MIF for women with metabolic syndrome was still clearly elevated when compared to healthy subjects (p=0.011). Circulating MIF concentrations showed a gender disparity between healthy and metabolic syndrome subjects. An elevation of systemic MIF in women with metabolic syndrome may contribute to pathogenesis of metabolic syndrome or to the development of metabolic syndrome-related diseases, such as atherosclerosis and type 2 diabetes mellitus.